Podcasts about k means

Link to bioRxiv paper: http://biorxiv.org/cgi/content/short/2023.04.09.536154v1?rss=1 Authors: Jajcay, N., Hlinka, J. Abstract: One of the interesting aspects of EEG data is the presence of temporally stable and spatially coherent patterns of activity, known as microstates, which have been linked to various cognitive and clinical phenomena. However, there is still no general agreement on the interpretation of microstate analysis. Various clustering algorithms have been used for microstate computation, and multiple studies suggest that the microstate time series may provide insight into the neural activity of the brain in the resting state. This study addresses two gaps in the literature. Firstly, by applying several state-of-the-art microstate algorithms to a large dataset of EEG recordings, we aim to characterise and describe various microstate algorithms. We demonstrate and discuss why the three "classically" used algorithms ((T)AAHC and modified K-Means) yield virtually the same results, while HMM algorithm generates the most dissimilar results. Secondly, we aim to test the hypothesis that dynamical microstate properties might be, to a large extent, determined by the linear characteristics of the underlying EEG signal, in particular, by the cross-covariance and autocorrelation structure of the EEG data. To this end, we generated a Fourier transform surrogate of the EEG signal to compare microstate properties. Here, we found that these are largely similar, thus hinting that microstate properties depend to a very high degree on the linear covariance and autocorrelation structure of the underlying EEG data. Finally, we treated the EEG data as a vector autoregression process, estimated its parameters, and generated surrogate stationary and linear data from fitted VAR. We observed that such a linear model generates microstates highly comparable to those estimated from real EEG data, supporting the conclusion that a linear EEG model can help with the methodological and clinical interpretation of both static and dynamic human brain microstate properties. Copy rights belong to original authors. Visit the link for more info Podcast created by Paper Player, LLC

Regression is a statistical and mathematical technique to find the relationship between two or more variables. In this episode of the AI Today podcast hosts Kathleen Walch and Ron Schmelzer define the terms Regression and Linear Regression and explain how they relate to AI and why it's important to know about them. Show Notes: FREE Intro to CPMAI mini course CPMAI Training and Certification AI Glossary AI Glossary Series – Machine Learning, Algorithm, Model Glossary Series: Machine Learning Approaches: Supervised Learning, Unsupervised Learning, Reinforcement Learning Glossary Series: Classification & Classifier, Binary Classifier, Multiclass Classifier, Decision Boundary Glossary Series: Clustering, Cluster Analysis, K-Means, Gaussian Mixture Model Continue reading AI Today Podcast: AI Glossary Series – Regression and Linear Regression at AI & Data Today.

ASHRAE Technical Editor Rebecca Matyasovski interviews Janice K. Means about her article in the April 2023 ASHRAE Journal. The article recognizes Eunice Newton Foote's discovery of the role CO2 plays in causing greenhouse effect. Means' new article series will spotlight historically undervalued contributions in areas that are of interest to ASHRAE and the HVAC&R industry.

The idea of grouping similar types of data together is the main idea behind clustering. Clustering supports the goals of Unsupervised Learning which is finding patterns in data without requiring labeled datasets. In this episode of the AI Today podcast hosts Kathleen Walch and Ron Schmelzer define the terms Clustering, Cluster Analysis, K-Means, and Gaussian Mixture Model, and explain how they relate to AI and why it's important to know about them. Continue reading AI Today Podcast: AI Glossary Series- Clustering, Cluster Analysis, K-Means, Gaussian Mixture Model at AI & Data Today.

In this episode of the AI Today podcast hosts Kathleen Walch and Ron Schmelzer define terms related to Machine Learning Approaches including Supervised Learning, Unsupervised Learning, Reinforcement Learning and explain how they relate to AI and why it's important to know about them. Show Notes: FREE Intro to CPMAI mini course CPMAI Training and Certification AI Glossary Glossary Series: Artificial Intelligence AI Glossary Series – Machine Learning, Algorithm, Model Glossary Series: Probabilistic & Deterministic Glossary Series: Classification & Classifier, Binary Classifier, Multiclass Classifier, Decision Boundary Glossary Series: Regression, Linear Regression Glossary Series: Clustering, Cluster Analysis, K-Means, Gaussian Mixture Model Glossary Series: Goal-Driven Systems & Roboadvisor Understanding the Goal-Driven Systems Pattern of AI Continue reading AI Today Podcast: AI Glossary – Machine Learning Approaches: Supervised Learning, Unsupervised Learning, Reinforcement Learning at AI & Data Today.

K-means clustering is an algorithm for partitioning data into multiple, non-overlapping buckets. For example, if you have a bunch of points in two-dimensional space, this algorithm can easily find concentrated clusters of points. To be honest, that's quite a simple task for humans. Just plot all the points on a piece of paper and find areas with higher density. For example, most of the points are located on the top-left of the plane, some at the bottom and a few at the centre-right. However, this is not that straightforward once you can no longer rely on graphical representation. For instance, when your data points live 3-, 4- or 100-dimensional space. Turns out, this is not that uncommon. Let me clarify. Read more: https://nurkiewicz.com/93 Get the new episode straight to your mailbox: https://nurkiewicz.com/newsletter

This week, please join Associate Editors Mercedes Carnethon and Karol Watson, as well as Guest Editor Fatima Rodriguez as they present the 2nd annual Disparities Issue. Then join Rishi Wadhera and Ashley Kyalwazi as they discuss their article "Disparities in Cardiovascular Mortality Between Black and White Adults in the United States, 1999 to 2019." Dr. Mercedes Carnethon: Well, good day listeners. I'm Mercedes Carnethon, and I'm joined by my fellow editors, Karol Watson, and Fatima Rodriguez, Associate Editor and Guest Editor for Circulation. And we'd like to welcome you to Circulation on the Run, for our second annual disparities issue. We have a lot of articles to discuss today, many of which we'll summarize, but we encourage you to access the issue and read the articles. First off, Fatima, I believe you have a paper to discuss. Dr. Fatima Rodriguez: Sure thing, Merci. My first paper is a thought provoking article by Nilay Shah, and co-authors from Northwestern University, that examine factors associated with the racial gap in premature cardiovascular disease. Dr. Fatima Rodriguez: This study used data from a well-known cardiac cohort, that aims to identify factors that begin in young adulthood and predict the development of future coronary artery risk. The objective of this study was to examine the relative contributions of clinical versus social factors, in explaining the persistent black/white gap in premature cardiovascular disease. After following around 5,000 black and white study participants for a median of 34 years, black men and women had a higher risk of premature cardiovascular disease. After controlling for multi-level individual and neighborhood level factors measured in young adulthood, the racial differences in premature cardiovascular disease were attenuated. Dr. Fatima Rodriguez: The authors found that the greater contributors to this racial disparity were not only clinical factors, but also neighborhood and socioeconomic factors. The relative explanatory power of each of these factors varied by men and women. This is really noteworthy, since we spent so much of our time in clinical medicine, focusing on identifying and managing traditional risk factors. But in reality, these structural factors and inequities are critically important to address, and contribute to differences in clinical risk factors downstream. Dr. Mercedes Carnethon: Thank you so much, Fatima. That was a really excellent summary. And now, I'm turning to you, Karol. I'd love to hear what you're going to be talking about today. Dr. Karol Watson: I'd like to discuss the paper, Association of Neighborhood Level Material Deprivation with Atrial Fibrillation Care in a Single-Payer Healthcare System Population Based Cohort Study. This is by Dr. Abdel-Qadir and colleagues. Dr. Karol Watson: So in this study, the author sought to determine whether there was an association between neighborhood material deprivation, by that we mean, inability to attain the basic needs of life and clinical outcomes, in individuals with atrial fibrillation. The kicker here is, they did this in an area with universal healthcare. So they wanted to see, if you took away the differences between the ability to see a physician or get your drugs paid for, if you would see any disparities. Dr. Karol Watson: So they performed a population based cohort study, individuals over the age of 66 years of age with atrial fibrillation, in the Canadian province of Ontario. They have universal healthcare there, and full drug coverage for anyone over 65. The primary exposure was neighborhood material deprivation. That's a metric used to estimate the inability to attain basic material needs, like healthy foods, safe housing. Neighborhoods were categorized by quintile, from the least deprived, quintile one, to the most deprived, quintile five. They find that, among about 350,000 individuals with atrial fibrillation, their mean age was 79, and about half of them were women. Those in the most deprived neighborhoods, quintile five, had a higher prevalence of cardiovascular risk factors and non-cardiovascular work comorbidity, relative to those who were in the least deprived areas. Dr. Karol Watson: Even after adjusting for all the confounders, they found that those in the most deprived neighborhoods had higher hazards of death, stroke, heart failure, and bleeding, relative to those in the least deprived neighborhoods. They also found that, despite having universal healthcare and drug coverage, those in the most deprived neighborhoods were less likely to visit a cardiologist, less likely to receive rhythm control intervention, such as ablation, and have worse outcomes. Dr. Karol Watson: And then, the accompanying editorial by Utibe Essien, he reminds us that intervening only on traditional markers of access, like health insurance and drug costs, may not be sufficient to achieve health equity. We have to address all of the structural needs that make people unable to get good help. Further, he points out that, the association between atrial fibrillation and neighborhood deprivation is very likely true with other cardiovascular conditions, as well. Dr. Karol Watson: So, Merci and Fatima, this just reminds us again, that addressing all the social determinants of health are necessary to achieve the best health outcomes. Dr. Mercedes Carnethon: Thanks so much, Karol. I really appreciate that summary of that important piece, focusing on a different domain of disparity. My first paper is an excellent piece, led by one of my favorite other associate editors at Circulation, Dr. Wendy Post, from Johns Hopkins University. And I see a familiar name on here. That's yours, Karol. You two are joined by an all-star list of authors, to describe race and ethnic differences in all-cause in cardiovascular mortality, in the multi-ethnic study of atherosclerosis. Dr. Mercedes Carnethon: MESA is a longitudinal cohort study that launched in 2000, and recruited just over 6,800 adults who identified as black, white, Hispanic, and Chinese. While the study participants were initially free from cardiovascular disease, over an average of 16 years of follow up, 364 participants died from cardiovascular disease. There are a number of novel findings in this paper that led our editor-in-chief to select it as his pick of the issue. Dr. Mercedes Carnethon: The finding that really stands out to me is, how much of an influence the social determinants of health had on black versus white disparities in cardiovascular mortality. In fact, after adjusting for socioeconomic status, the disparities were nearly eliminated. Other critically important findings are that, the oft described Hispanic paradox of lower cardiovascular mortality in Hispanics, as compared with white adults, was demonstrated in this population. And finally, we have longitudinal data on Asians living in the United States. Asian participants in MESA had similar rates of cardiovascular disease mortality as their white counterparts. There's so much to learn in this well designed cohort study, and so many hypotheses about how social determinants and structural racism influence the disparities that we see. Dr. Mercedes Carnethon: So Fatima, I'd like to turn to you next. What else do you have to share? Dr. Fatima Rodriguez: Thank you, Merci. My second paper is a research letter for my home institution of Stanford University, led by my colleague, Dr. Shoa Clarke, discussing how race and ethnicity stratification for polygenic risk course, may mask disparities among Hispanic individuals. Dr. Fatima Rodriguez: This study used data from the Million Veteran Program, to determine how self-identified race and ethnicity impact the performance of polygenic risk scores in predicting coronary artery disease. Dr. Fatima Rodriguez: The investigators found, that the current polygenic risk scores predict coronary artery disease similarly well in Hispanic and non-Hispanic white individuals. However, what I found most interesting, is that there was so much more heterogeneity among Hispanic individuals as measured by K-Means clustering, than among non-Hispanic white individuals. And this study really confirms that there is much more heterogeneity within populations than between populations. And this is particularly true as we think of the extreme diversity of Hispanic populations. Lumping Hispanic populations into one category, may mask important differences in cardiovascular risk prediction outcomes, and even the notions of the Hispanic paradox that you just discussed, Merci. Dr. Mercedes Carnethon: I appreciate you bringing that up again, because there are so many different nuances to the observations that we see in these studies. But I'll keep moving, because we have an embarrassment of riches in this wonderful issue. So Karol I'll turn back to you. Dr. Karol Watson: Thanks, Merci. The next paper I'd like to discuss, is an On My Mind piece by Peter Liu and colleagues, and they entitle it, Achieving Health Equities in the Indigenous Peoples of Canada, Learnings Adaptable for Diverse Populations. Now the author's note that, lessons learned about addressing health disparities from indigenous peoples in Canada, can offer a lot of new lessons for other populations where there are similar disparities. They begin by offering historical perspective, and they say that, most of the health to disparities for the indigenous populations originate from early colonization, in dismantling of the sociocultural economic educational and health foundations, the indigenous communities had historically. Dr. Karol Watson: It's true that, that is true in a number of different countries. This is data from Canada, but we can see similar things in the United States. With the recognition of the historical and ongoing social health inequities, the Canadian government initiated what they call, the Truth and Reconciliation Commission, to recommend a path towards reconciliation, to create best practices for engaging indigenous populations. Dr. Karol Watson: For instance, in Canada, any health research or implementation program, requires the direct engagement of indigenous communities and their elders. They have to try to develop culturally safe environment, including what they say, quote unquote, anti-racism and cultural safety education for all, both indigenous and non-indigenous populations. They want to really respect community values, customs and traditions, including the access to traditional foods, and healing practices, and the support from elders. So they really are making it a very important point, that cultural sensitivity is absolutely critical to engaging these populations. You want to jointly collect data whenever available, to track progress and outcomes. And they offer many examples of successful programs developed using these principles, such as the Diabetes and My Nation program, in British Columbia, or the mobile diabetic telehealth clinic. Dr. Karol Watson: They offer discussion of future initiatives as well, that can help other communities in Canada. Such as, there's an initiative addressing hypertension in the Chinese population in Canada. Dr. Karol Watson: So this thoughtful paper, really looks at disparities in unique populations in Canada. More importantly, it offers potential roadmaps for other populations, solutions to address longstanding legacies of racism and colonialism. Dr. Mercedes Carnethon: Thank you so much, Karol, for that description from our neighbors from the north. Dr. Mercedes Carnethon: My second paper is really relevant during this hot month of July, in much of the United States and the upper hemisphere. And that's because Sameed Khatana and colleagues from the University of Pennsylvania, discuss how extreme heat is associated with higher cardiovascular mortality. For those of us who welcome the heat of summer and the opportunity to get out from behind our desks and exposed to some vitamin D, Khatana and colleagues reviewed county level daily data on temperature, and linked those data with mortality rates. Dr. Mercedes Carnethon: But before I summarize the findings, I invite you to California based cardiologists to join me in Chicago, where extreme heat is really only a problem for about 30 days a year. The authors found that between 2008 and 2017, when the heat index was above 90 degrees Fahrenheit, or 32.2 degrees Celsius, there was a significantly higher monthly cardiovascular mortality rate. In total, extreme heat was associated with nearly 6,000 additional deaths from cardiovascular disease. And sadly, black adults, older adults, and men, bore the greatest burden of mortality rates from extreme heat. So, we can all take lessons from that. Dr. Mercedes Carnethon: But turning to you now, Fatima. Dr. Fatima Rodriguez: Thanks so much, Merci. I'm from Florida, so I can definitely relate to the issues of extreme heat, but I'm very happy for the perfect year round weather here in Northern California. Dr. Fatima Rodriguez: My third paper is led by Dr. Zubair (and Chikwe) and colleagues from Cedar Sinai, and it describes changes in outcomes by race, in children listed for heart transplantation in the United States. I won't give all the details, but this research letter really nicely summarizes how the 2016 Pediatric Heart Allocation Policy revisions may have inadvertently widened health disparities between white and non-white children. This article touches on the difference between equality and equity, even in the most well-intentioned national policies. And I invite our listeners to read the full details in this special Circulation edition. Dr. Mercedes Carnethon: Thanks Fatima. Karol. Dr. Karol Watson: The next paper I'd like to discuss, is a community based cluster randomized pilot trial, of a cardiovascular mobile health intervention, preliminary findings of the FAITH! Trial, from LaPrincess Brewer and colleagues from the Mayo Clinic. Dr. Karol Watson: So it's well known that African Americans have suboptimal cardiovascular health metrics, such as less regular physical activity, suboptimal blood pressure levels, suboptimal diets. So the authors of this study hypothesize, that developing a mobile health intervention, in partnership with trusted institutions, such as, African American churches, might be an effective means to promote cardiovascular health in African American patients. So using a community based participatory research approach, they develop the FAITH! trial. FAITH stands for Fostering African American Improvement in Total Cardiovascular Health. The manuscript in this issue reports, feasibility and preliminary efficacy findings from this refined community informed mobile health intervention, using the FAITH! App, developed by the investigators. Dr. Karol Watson: They performed a cluster randomized control trial. Participants from 16 different churches in the Rochester, Minnesota and Minneapolis St. Paul, Minnesota areas. The clusters were randomized to receive the FAITH! App, that was the intervention group, or were assigned to a delayed intervention program. The 10 week intervention feature culturally relative and sensitive information modules, focused on American Heart Association's Life's Simple 7. Primary outcomes were changes in the mean Life Simple 7 score, from baseline to six months post intervention. They enrolled 85 participants, mean age was 52, and about 71% were female. Dr. Karol Watson: At baseline, the mean Life Simple 7 score was 6.8, and 44% of the individuals were characterized as being in poor cardiovascular health. The mean Life Simple 7 score of the intervention group, after the end of the intervention, increased by 1.9 points. In the control comparator group, it only increased by 0.7 point. Highly statistically significant, with P value of less than 0.0001 at six months. Dr. Karol Watson: Now this FAITH! Trial demonstrated preliminary findings, that suggest that a culturally sensitive and mobile health lifestyle intervention could be efficacious, promoting ideal cardiovascular health among African Americans. I think what's so important about this is that, they partnered with a very trusted group, the churches, and getting buy-in to a community that has had many reasons not to trust in the past, I think is critically important. Dr. Mercedes Carnethon: Well, thank you so much, Karol. My third paper is an original research investigation by Anoop Shah and colleagues from the University of Edinburgh, arguing that socioeconomic deprivation is an unrecognized risk factor for cardiovascular disease. Dr. Mercedes Carnethon: In their study, the authors evaluated how risk scores, with and without indicators of socioeconomic deprivation, performed in a population study in Scotland, the Generation Scotland: the Scottish Family Health Study, of over 15,000 adults. Again, I won't give away all the details, so that I keep our listeners excited to read the article, but all risk scores aren't created equally. And the observed versus expected number of events varied, based on whether the risk score included socioeconomic indicators or not. Further, the performance of the risk scores varied, based on the degree of deprivation that participants were currently experiencing. It's a thought provoking piece, that may challenge us to reconsider how we identify risks for cardiovascular disease in the population. Dr. Mercedes Carnethon: And I'm turning to you now, Fatima. Dr. Fatima Rodriguez: Sure thing, Merci. My last paper is led by Dr. Anna Krawisz, and is looking at how differences in comorbidities explain racial disparities in peripheral vascular interventions. This study used Medicare fee for service data from 2016 to 2018, to examine risks of death and major amputation, one year following peripheral endovascular intervention. They found that, black Medicare beneficiaries had higher population level need for peripheral endovascular interventions, and that black race was associated with adverse events following these interventions. However, after adjusting for the higher prevalence of comorbidity, such as diabetes, hypertension, and chronic kidney disease in black populations, this observation was eliminated. Again, like a common theme in many of the articles we've discussed today, this is to suggest, that moving upstream to reduce risk factors is really critical to eliminate disparities in cardiovascular disease outcomes. And this includes the understudy disease of peripheral arterial disease. Black adults were also less likely to be treated with guideline directed medical therapies in this study. Dr. Mercedes Carnethon: Well, thank you so much, Karol and Fatima, for your wonderful summaries of all of the excellent pieces in this issue. Dr. Karol Watson: And I'd like to thank all of the fantastic investigators who submitted their really fantastic work, so that we could produce this issue. And really, keep them coming. We thank you for this. Dr. Mercedes Carnethon: Well, thank you. So now we'll transition to our feature discussion with Drs. Wadhera and Kyalwazi, from Beth Israel Deaconess Medical Center, and the Harvard Medical School. Dr. Mercedes Carnethon: Welcome to this episode of Circulation on the Run podcast. I'm really pleased to host this feature discussion. My name is Mercedes Carnethon, from the Northwestern University Feinberg School of Medicine. And I'm pleased to have with us today, Drs. Ashley Kyalwazi and Rishi Wadhera from Beth Israel Deaconess, and the Harvard Medical School. And they shared with us a really important piece of work for our disparities issue, that is describing disparities in cardiovascular mortality, between black and white adults in the United States from 1999 to 2019. First of all, I really want to thank you both for submitting your important work to circulation. Dr. Rishi Wadhera: Thanks so much Mercedes, and thanks for the opportunity to submit and revise our manuscript. Ms. Ashley Kyalwazi: Thanks so much for having us. Dr. Mercedes Carnethon: Wonderful. I'd like to start out with you Rishi. Tell our listeners about the objectives of your study, and what your motivation was for carrying out this work. Dr. Rishi Wadhera: Well, I think it's been well established that, black adults are disproportionately impacted by cardiovascular disease, and experience worse cardiovascular outcomes, due to systemic inequities and structural racism. And so, the goal of our study was really, to perform a comprehensive national evaluation of how age adjusted cardiovascular mortality rates have changed for black adults, compared with white adults, over the past two decades in the United States, with a focus on some key subgroups, like younger adults and women. Dr. Rishi Wadhera: In addition, because we know that the neighborhood community or environment in which you live in the US, has an immense influence on cardiovascular health, we also examine changes in cardiovascular mortality for black and white adults by geographic region, rurality, and neighborhood racial segregation. And our primary objective was really, to understand whether disparities in cardiovascular outcomes between black and white adults improved, worsened, or didn't change, from 1999 to 2019. Dr. Rishi Wadhera: And there are some reasons to think we might have made progress in narrowing the mortality gap between these groups over this time period. There have been substantial improvements in preventative care and treatments for cardiovascular disease over the past two decades. And the expansion of insurance coverage under the Affordable Care Act, led to increases in access to care, cardiovascular risk factor screening and treatment, particularly, for black adults. At the same time, we know that, black adults were disproportionately affected by the economic recession of 2008, and experienced worsening poverty, job loss, and wealth loss, all of which are inextricably tied to cardiovascular health, and more broadly, health. And so that was our interest in really exploring how disparities in cardiovascular mortality have changed amongst black and white adults between 1999 and 2019. Dr. Mercedes Carnethon: Thank you so much for that summary. It's really nice to have these sort of pieces that really outline for us a lot of data, and across a number of different domains. Because it allows us really, a chance to think about those data, and how we can use those data in order to help improve health. Dr. Mercedes Carnethon: So tell me a little bit, Ashley, about what your study found. Ms. Ashley Kyalwazi: Absolutely. Yeah. So in the United States, overall, we found that age adjusted cardiovascular mortality rates declined for both populations, so both black and white adults, by around 40% from 1999 to 2019. So encouraging declines across the country. We found that these patterns were similar for both women and men, when we stratified by gender, over the 20 year period. While mortality rates declined in all regions, we still did find disparities when we stratified by age. So between the younger and older black women, versus younger and older black men, we found that, younger black men and black women were dying at higher rates, and were at increased risk of death from cardiovascular mortality, compared to younger white women and men, respectively. But we also found that black women and men living in rural areas consistently experienced highest mortality rates. And then finally, black adults living in higher areas of residential racial segregation, and compared to those living in low to moderate areas of residential racial segregation had higher mortality rates, as well. Dr. Mercedes Carnethon: Wow, this is a lot. And it's really describing a lot of disparities across multiple domains that we can easily measure. Which aspects of these results in your work did you find the most surprising, Ashley? Ms. Ashley Kyalwazi: Yeah, I was intrigued, I think overall, by just the gaps. I was very encouraged by, I think, the declines over time. On an absolute scale, the country has made a lot of progress, in terms of reducing cardiovascular mortality rates for both groups. But still, by the end of the study period, there were notable gaps between black adults and white adults. Particularly, between black, younger women and white, younger women, we see that by the end of the study period, black, younger women still remain over two times the risk of death from cardiovascular disease than younger white women. Which I think, leaves something to be desired from a public health and health policy standpoint, with regards to how we're going to kind of decrease these disparities. Dr. Mercedes Carnethon: I wanted to follow up on that point. Why do you think you see such disparities between black and white younger women? I love the opportunity of the podcast to allow authors a chance to speculate, beyond what they would do in the paper. Ms. Ashley Kyalwazi: Absolutely. I think that, there are a lot of great efforts on a national scale right now, to kind of address the disparities between black and white women. The Association of Black Cardiologists, for example, had a whole paper out about ways to really target and provide preventative measures for black women. So for example, working with communities, where there's a high proportion of black women, to figure out what community based research looks like. Engaging with churches, different types of methods, to really understand the barriers that black women face towards obtaining preventative care. Ms. Ashley Kyalwazi: I think the disparities that we are seeing, could potentially parallel well known and documented disparities in maternal health outcomes. So I think, from a perspective of preventative care, really thinking about, what are the barriers to healthy cardiovascular profiles for black women pre and postnatally, to ensure that their cardiovascular health is an actionable before and after the pregnancy? Ms. Ashley Kyalwazi: And then I think, broadly, the challenges that black women face, mirror the challenges of black adults, plus the additions of like social stressors, things that we looked at in this study neighborhood residential racial segregation, access to healthcare, and all of those things kind of contribute to the profile that black women face, in terms of being often, the heads of their households as well, and carrying on a lot of different societal challenges. Dr. Mercedes Carnethon: Thank you so much for that. I really appreciate that. Dr. Mercedes Carnethon: As I read the paper, one of the findings that I found the most surprising, and it was challenging for me to understand, is that while the absolute difference in rates was declining, or getting smaller over time, between black and white men and women, the rate ratios remained elevated across the course of time. I think, these concepts can be a little challenging to understand, not just to me, but to others as well. That when one measure of effect is showing progress, but another is still reporting a disparity. Dr. Mercedes Carnethon: Rishi, could you explain for our listeners, how we can see progress on one metric, but still find a mortality rate ratio that's 1.3 times higher in black, as compared with white men, for example? Dr. Rishi Wadhera: Thanks for that really important question, Mercedes. Just to summarize, we presented two outcomes that compared cardiovascular deaths among black and white adults in our paper, absolute rate differences, and then separately, rate ratios. And I think, both measures provide important complementary insights. I think that, understanding the absolute rate difference in cardiovascular deaths is critically important from a public health perspective, because it characterizes excess deaths experienced by black adults, compared with white adults. The fact that the absolute rate difference in cardiovascular death has narrowed over the past two decades between these groups is positive news. In contrast, the rate ratio provides us with important insights on the relative difference, or disparity or gap, between black and white adults. Dr. Rishi Wadhera: So again, both are important, both provide sort of synergistic and complimentary insights. And just to sort of cement that, as an example, you were talking to Ashley earlier, about some of the patterns we noticed amongst younger black women and white women. The absolute rate difference in cardiovascular deaths between younger black women, compared to younger white women, decrease from 91 per 100,000 in 1999, to about 56 per 100,000 in 2019. And that's good progress. However, our rate ratio analysis indicated that, still in 2019, young black women were 2.3 times more likely to die of cardiovascular causes than young white women. Highlighting that, we still have a lot of work to do, to address disparities between these groups. Some of which, Ashley already talked about. Dr. Mercedes Carnethon: Thank you so much for that excellent explanation. I know it's certainly, I find it alarming to hear, but then I remember I'm actually not young anymore. So maybe this doesn't apply to me quite as much. But no, I appreciate the explanation. Dr. Mercedes Carnethon: So your report was really unique, in that you studied these disparities, as we discussed, across a number of domains, age, geography, even racial residential segregation. Whereas, the pronounced disparities have been reported in a few of the other domains that you studied. I'm really interested in hearing more about racial residential segregation. I think, a lot of people don't fully understand what the concept is, and the ways in which racial residential segregation may contribute to higher rates of cardiovascular death among blacks. Dr. Mercedes Carnethon: Ashley, would you mind explaining to us first, what racial residential segregation is? And then really, how it would contribute to higher rates of cardiovascular death? Ms. Ashley Kyalwazi: Yeah, absolutely. So in its simplest terms, racial residential segregation is just the physical separation of two or more groups by race and/or ethnicity into different neighborhoods. What gets tricky is, like the long history within the United States of how we got to this point, where you see numerous degrees of segregation across the country. Residential racial segregation in the United States dates back to policies pre World War II, that resulted in kind of discriminatory banking practices and policies. For example, reverse red lining and gentrification, much of which the extent still exists today. And that's what we see kind of, I think, in our results when we looked at high versus low to moderate areas of residential racial segregation, and how those kind of track onto the trends that we see in cardiovascular mortality over time. Ms. Ashley Kyalwazi: The residential racial segregation impacts almost every aspect of life. You can imagine where you live, we know definitely impact, for example, your zip code can impact health outcomes. We've seen individual's cardiovascular health kind of trend with something as simple as your zip code. Where you live really does impact your, for example, access to affordable housing, health insurance, where your primary care physician is, whether or not you even have one. What that trip looks like to see your primary care physician, is it hours on end, and unrealistic to get to, or is it just around the corner? Ms. Ashley Kyalwazi: Educational opportunities, which leads to income, which we know is linked to cardiovascular disease employment in all of these aspects. Even access to green space. In some metropolitan areas that are more segregated, we see that, black adults, for example, have less access to green space, and numerous studies have shown that, that does impact overall health, but then also, from a cardiovascular disease perspective as well. So I think that, given that we know that lack of access to all of these key determinants can adversely affect cardiovascular mortality, and just general cardiovascular health, I think is very interesting that we found that, there was this link between high residential racial segregation and cardiovascular mortality. That we definitely can look into more, and understand kind of in more detail, that the mechanisms at play and ways to intervene. Dr. Rishi Wadhera: And just to layer onto and reinforce Ashley's really excellent answer to that question. We know that black adults are more likely to live in disadvantaged neighborhoods, because of the intentionally racist policies that were put in place many decades ago, that Ashley described so well. And black communities and segregated communities, as Ashley mentioned, are less likely to have access to primary care, high quality hospital care, and green spaces, but also, pharmacies and healthy foods. And we also know, there's a lot of empirical work that's shown that black communities, disproportionately experience psychosocial stressors, trauma. Dr. Rishi Wadhera: Also, these communities are disproportionately exposed to climate change, such as extreme heat. There was a recent paper that extreme heat has been linked to increases in cardiovascular mortality, and disproportionately affects black communities. These communities are also disproportionately exposed to pollution. All of these things we know are linked to cardiovascular health, and represent the effects of again, intentionally racist policies that were put into place many decades ago, the effects of which still persists today. Which will require equally intentional policies that aim to dismantle these longstanding effects, if we hope to make progress in advancing health equity, and specifically, cardiovascular health equity. Dr. Mercedes Carnethon: I appreciate the facility with which the two of you address the multiple complex contributors to cardiovascular health. It's even more impressive coming from two clinicians. So I really appreciate you taking the time to explain this. And this is where I really like the opportunity to open up and say, what more do you want your clinical peers to know about? For example, how does this affect the day to day encounters that you have in clinic with black patients, and other patients who've been traditionally underrepresented? How do you hope your clinical peers will use this information to promote cardiovascular health equity? And I'll open it up to either of you to respond. Ms. Ashley Kyalwazi: Yeah, I can get on that one. I think that, the disparities that our paper highlights, really requires a multisystem level approach to tackling, from public health to public policy. But I think at a provider level, to your question, Mercedes, physicians must be able to, I think at first, read the data and understand that these disparities exist. Ms. Ashley Kyalwazi: If there's no insight with regards to the risk profiles, that simply black women and black men have, because of systemic racism, because of these inequities, then I think, we're already kind of steps behind where we need to be. So recognizing disparities in cardiovascular disease burden for black men and women, prioritizing education on cardiovascular risk. A lot of the conditions are preventable with appropriate access to care and education around these topics. And so, providing education about the signs and symptoms of heart disease and treatment options for black men and women. Recognizing the history of medical mistreatment for black adults in this country. And really, tailoring the approach towards the individual who comes into the office, who might have very valid reasons for hesitating to take a medication, or a lot of questions that need time and consideration. Ms. Ashley Kyalwazi: At a research level, I think, more data and resources should be spent on studying risk prevention and treatment for cardiovascular disease in black adults, and really, developing more community based models, that really get at the specific interventions that work within black communities, that are culturally specific, that are targeted and relevant, for the populations that we're talking about. Ms. Ashley Kyalwazi: I think finally, and I'll let Rishi chime in, I think, this is shockingly low level of racial and ethnic representation in the field of cardiology as a whole. And we know that, diversity in healthcare can improve health outcomes. So from a cardiology perspective, I think, training the next generation of black young men and women to take up their seats at the table, and really advocate for some of these issues, alongside individuals who are already doing great work, would be essential towards reducing disparities that we see. And so all of the above, I think, I would encourage for my colleagues. Dr. Mercedes Carnethon: Thank you so much. Rishi, any final thoughts? Dr. Rishi Wadhera: No, I'll just add onto Ashley's again, really outstanding response that, this is a tension we face when we see patients in cardiology clinic all the time. I think, awareness about disparities, and the multiple factors that contribute to disparities in cardiovascular health, particularly, as it relates to race and ethnicity, are increasingly being recognized as they should be. Dr. Rishi Wadhera: And one of the challenges, how much can clinicians do within the bounds of hospital walls? We can make sure that we get patients the treatments they need. We can make sure we screen patients appropriately. But we know, as we've discussed, that so many factors beyond hospital walls, like widening income inequality, that's disproportionately affected black adults, and has been worsening over the last several decades. Widening educational inequality, that again, disproportionately affects black adults, and has been worsening over decades, also affect how. So I think, thinking about how clinicians, researchers, and policy makers, can work together to address some of these challenges, issues, and broader social determinants of health, that also exist outside our clinical practice, or hospital walls, will be really, really important, if we are serious about advancing health equity in this country. Dr. Rishi Wadhera: I don't think, we can operate in silos anymore. In the clinical world, in the research world, in the policy making world, we need more researchers and clinicians to have a seat at the table when it comes to policy making, individuals who understand how all of these complex factors are inextricably tied to one another, so that we can seek and implement solutions that advance cardiovascular health. Dr. Mercedes Carnethon: Thank you so much. The insights that we've gotten, from not only your written work, but even more importantly, this opportunity to speak with you today, and share with our readership, have just been invaluable. And I really appreciate the amount of time that you spent, in preparing the manuscript, and really contextualizing the findings with us today, as well as in writing. So thank you so much for contributing this really important work to our annual disparities issue. Dr. Rishi Wadhera: Thank you so much, Mercedes. We really appreciate all the time you and the Circulation team took to make the manuscript stronger. Ms. Ashley Kyalwazi: Thank you so much for having us. It was truly an honor to have this conversation and to submit our work. Dr. Mercedes Carnethon: Well, thank you. Dr. Mercedes Carnethon: That wraps up our feature discussion for this episode of Circulation on the Run podcast. I'm Mercedes Carnethon, from Northwestern University, Associate Editor and guest editor of the disparities issues. So thank you so much. Dr. Greg Hundley: This program is copyright of the American Heart Association, 2022. The opinions expressed by speakers in this podcast are their own, and not necessarily those of the editors, or of the American Heart Association. For more, please visit hajournals.org.

Cet épisode marathon sera découpé en deux morceaux pour éviter à vos oreilles une écoute marathon. Dans cet épisode on y parle Brian Goetz, Bian Goetz, Brian Goetz, usages des threads virtuels, OpenAPI, Kubernetes, KNative, copilot et Tekton. La deuxième partie couvrira des sujets d'architecture et de loi société et organisation ainsi que les conférences à venir. Enregistré le 8 juillet 2022 Téléchargement de l'épisode LesCastCodeurs-Episode–281.mp3 News Langages Peut-être une nouvelle syntaxe spécifique aux Records Java pour tordre le cou aux builders Brian Goetz discute de l'idée d'avoir une syntaxe spécifique pour les records pour facilement créer un record dérivé, potentiellement avec des valeurs par défaut, mais en paramétrant certains champs Point shadowPos = shape.position() with { x = 0 } Cela évite de créer la notion de paramètre par défaut dans les constructeurs ou les méthodes Il y a l'article Data Oriented Programming de Brian Goetz, sur InfoQ projet Amber amène des changements qui combinés permet de faire du data oriented programming en Java et pas que du OOP OO combine état et comportement (code) OO est super utile pour défendre des limites (programme large en des limites plus petites et plus gérable) mais on s'oriente vers des applications plus petites (microservices) data oriented programming: modélise data immuable et le code de la logique métier est séparée records -> data en tant que classe, sealed classes -> définir des choix, pattern matching -> raisonne sur des data polymorphiques algebraic data: hiérarchie de sealed classes dont les feuilles sont des records: nommées, immuable, testable (pas de code) Un nouveau JEP pour intégrer une Classfile API Le JDK inclut déjà des forks de ASM, de BCEL, et d'autres APIs internes, pour manipuler / produire / lire le bytecode Mais l'idée ici c'est que le JDK vienne avec sa propre API officielle, et qui soit plus sympa à utiliser aussi que le pattern visiteur de ASM par exemple La version d'ASM intégrée était toujours en retard d'une version (problème de poule et d'oeuf, car ASM doit supporter la dernière version de Java, mais Java n+1 n'est pas encore sorti) Lilian nous montre à quoi va ressembler les Record Patterns de JEP 405 Apache Groovy et les virtual threads, et aussi Groovy et le Deep Learning Paul King, qui dirige actuellement le PMC de Apache Groovy, a partagé récemment plusieurs articles sur le blog d'Apache sur des intégrations intéressantes avec Groovy Groovy et sa librairie GPars pour la programmation concurrente et parallèle s'intègre facilement avec les Virtual Threads de JEP 425 / JDK 19 https://blogs.apache.org/groovy/entry/gpars-meets-virtual-threads Groovy avec Apache Wayang et Apache Spark pour classifier des Whiskey par clusterisation KMeans https://blogs.apache.org/groovy/entry/using-groovy-with-apache-wayang Et aussi Groovy avec différentes librairies de Deep Learning pour la classification https://blogs.apache.org/groovy/entry/classifying-iris-flowers-with-deep Le jargon (en anglais) de la programmation fonctionnelle, si vous avez rêvé d'avoir sous la main la définition de foncteur, de monoïde, et j'en passe avec des exemples en JavaScript des pointeurs vers des librairies fonctionnelles en JavaScript des traductions dans d'autres langues et d'autres langages de programmation Librairies Spring Boot 2.7 SpringBoot 2.7 Spring GraphQL 1.0 Support pour Podman Gestion de dépendance et auto configuration pour Cache2k nouvelle annotations pour Elasticsearch et CouchBase dernière versions avant SpringBoot 3 qui changera plus de choses. Recommande de migrer une version a la fois. Support pour 2.5 à fini (upstream) Quarkus 2.10.0 Travaux préliminaires sur les threads virtuels de Loom Support non-blocking pour GraphQL Prise en charge des Kubernetes service binding pour les clients SQL réactifs CacheKeyGenerator pour l'extension de cache quarkus-bootstrap-maven-plugin déprécié et remplacé par quarkus-extension-maven-plugin (uniquement utile pour les développeurs d'extensions Quarkus) Nouveaux guides: Using Stork with Kubernetes OpenId Connect Client Reference Guide Using Podman with Quarkus Les différences entre OpenAPI 2 et 3 Introduction de la notion de lien pour créer des relations entre Response et Operations, pratique pour faire des APIs hypermédia La structure du document OpenAPI a été -un peu simplifiée, en combinant par exemple basePath et schemes, ou en rassemblant les securityDefinitions Des améliorations sur les security schemes, autour de OAuth et OpenID Plus de clarté dans la négociation de contenu et les cookies La section des exemples de Request / Response devrait aider les outils qui génèrent par exemple des SDK automatiquement à partir de la description OpenAPI Un support étendu de JSON Schema Introduction d'une notion de Callback, importante pour les APIs asynchrones, en particulier les WebHooks je me demande si ils ont l'intention d'embrasser AsyncAPI ou su la partie asynchrone d'OpenAPI 3 a pour objectif de faire de la competition Infrastructure N'utilisez pas Kubernetes tout de suite ! Kubernetes, c'est bien, mais c'est un gros marteau. Est-ce que vous avez des gros clous à enfoncer ? Ne commencez peut-être pas avec l'artillerie lourde de Kubernetes. Commencez plutôt avec des solutions managées genre serverless, ce sera plus simple, et au fur et à mesure si votre infrastructure a besoin de grossir et dépasse les fonctionnalités des solutions managées, à ce moment là seulement évaluer si Kubernetes peut répondre à votre besoin Choisir Kubernetes, c'est aussi avoir la taille de l'équipe qui va bien avec, et il faut des profils DevOps, SRE, etc, pour gérer un cluster K8S L'auteur suggère grosso modo que ça dépend de l'ordre de magnitude de la taille de l'équipe : avec quelques personnes, préférez des solutions type Google App Engine ou AWS App Runner, avec une dizaine de personne peut-être du Google Cloud Run ou AWS Fargate, avec moins d'une centaine là pourquoi pas du Kubernetes managé comme Google Kubernetes Engine, et si vous dépassez mille, alors peut-être vos propres clusters managés par vos soins et hébergés par vos soins sur votre infra ca impose d'utiliser les services du cloud provider? Parce que la vie ce n'est pas que du code maison. C'est la mode de dire de pas utiliser K8S : https://www.jeremybrown.tech/8-kubernetes-is-a-red-flag-signalling-premature-optimisation/ (mais bon, vu le nombre de fois où il est pas utilisé à b Knative Eventing Devlivery methods on peut faire de la delviery simple 1–1 sans garantie on peut faire de la delivery complexe et persistante en introduisant la notion de channel qui decouple la source de la destination. on peut repondre a la reception d'un message et pousser la réponse dans un second channel mais ca devient compliquer a gérer quand on rajoute des souscripteurs il y a la notiuon de broker qui definit: des flitres, un channel (automatique) et la capacité de répondre les triggers sont un abonnement non pas a un channel mais a un type d'évènement spécifique Cloud AWS is Windows and Kube is Linux pourquoi utilisez Kube qui etait pas stablewa lors qu'AWS offre tout AWS forcé d'offrir EKS MAis pourri Lockin AWSIAM Pourquoi AWS serait le windows economies d'echelles de faire chez soi kube devient rentable une certaine taille de l'organisation besoin alternative a AWS (bus factor) on voit le Kube distro modele arriver Google data center Paris Outillage IntelliJ IDEA 2022.5 EAP 5 amène des nouveautés Frameworks and Technologies Spring 6 and Spring Boot 3 Support for new declarative HTTP Clients in Spring 6 URL completion and navigation for Spring Cloud Gateway routes Experimental GraalVM Native Debugger for Java Code insight improvements for JVM microservices test and mock frameworks Code insight improvements for Spring Shell Improved support for JAX-RS endpoints Support for WebSockets endpoints in HTTP Client Support for GraphQL endpoints in the HTTP Client UI/UX improvements for the HTTP Client Improved navigation between Protobuf and Java sources Kubernetes and Docker Intercept Kubernetes service requests with Telepresence integration Upload local Docker image to Minikube and other connections Docker auto-connection at IDE restart Docker connection options for different docker daemons GitHub copilot est disponible pour tous (les developpeurs) 40% du code écrit est généré par copilot en python (ca calme) gratuit pour les étudiants et les développeurs OSS Revue de Redmonk décrit copilot comme une extension d'intelligence ou auto complete mais qui « comprend » le code autour premiere fois pas une boite de cette taille et à cette échelle l'avantage de copilot en terme de productivité, de qualité de code, de sécurité et de légalité En gros, c'est encore à voir. Mais la qualité impressionne les gens qui l'ont testé ; sécurité pas de retour d'un côté ou de l'autre sauf que les développeurs humains ne sont pas des lumières de sécurité :D GitHub pense que GitHub n'est pas responsable de la violation de code vue que ce sont des machines et des algorithmes qui transforment: cela a l'air d'etre le consensus des avocats GitHub dit qu'on est responsable du code qu'on écrit avec copilot Et implicitement GitHub dit que la licensure du code « source » ne se propage pas au code generé. Et là, c'est pas clair et de la responsibilité de l'utilisateur, mais la encore les avocats sont plutot ok moralement c'est probablement pas ok mais bon et il y a débat autour des licenses copyleft notamment LGPL 1% du temps, code copié verbatim de > 150 caractères Question sur le code non open source sur lequel GitHub Copilot s'appuie mais en gros le marcher s'en fout un peu des licences Risque de reputation de Microsoft la question c'est quand / si les gens seront prêt à accepter cet usage Gradle publie sa roadmap Historiquement, la société Gradle Inc ne publiait pas vraiment de roadmap officielle Outre les tickets que l'on pouvait voir dans Github, cette fois ci, une “roadmap board” est visible et disponible pour tout le monde, et pas seulement pour les clients Tekton est groovy (mais non, il n'utilise pas Groovy !) Un grand tutoriel sur Tekton Une brève histoire de CI/CD (avec un contraste avec Groovy utilisé dans Jenkins) Un aperçu des grands concepts de Tekton, avec ses tâches et ses pipelines (Task, TaskRun, Pipeline, PipelineRun) Comment installer Tekton Les outils CLI Un exemple concret d'utilisation Sortie de Vim 9, surtout avec VimScript 9 des changements incompatibles entre VimScript 8.2 et 9 font qu'il était nécessaire de passer à une version majeure mais l'ancienne version du langage reste supportée pour compatibilité avec la nouvelle, les utilisateurs peuvent s'attendre à des performances x10 voire x100 ! le langage devient pré-compilé, au lieu d'être interprété ligne par ligne l'idée était d'avoir un langage plus proche de ce qu'on trouve dans JavaScript, TypeScript ou Java Conférences De la part de Youen Cette année Codeurs en Seine, c'est le 17 novembre et le cfp est ouvert N'hésitez pas à amener un peu de JVM dans l'appel à orateur. (ca commence à se faire rare). Pour rappel : codeurs en seine c'est 1000 personnes autour des métiers du développement dans une des plus grande salle de Rouen, le kindarena. Nous contacter Soutenez Les Cast Codeurs sur Patreon https://www.patreon.com/LesCastCodeurs Faire un crowdcast ou une crowdquestion Contactez-nous via twitter https://twitter.com/lescastcodeurs sur le groupe Google https://groups.google.com/group/lescastcodeurs ou sur le site web https://lescastcodeurs.com/

In this episode, we interview Jonas Landman, a Postdoc candidate at the University of Edinburg. Jonas discusses his study around quantum learning where he attempted to recreate the conventional k-means clustering algorithm and spectral clustering algorithm using quantum computing.  Click here to access additional show notes on our website!

K-means is widely used in real-life business problems. In this episode, Mujtaba Anwer, a researcher and Data Scientist walks us through some use cases of k-means. He also spoke extensively on how to prepare your data for clustering, find the best number of clusters to use, and turn the ‘abstract' result into real business value. Listen to learn.  Click here to access additional show notes on our website! Thanks to our sponsor! ClearML is an open-source MLOps solution users love to customize, helping you easily Track, Orchestrate, and Automate ML workflows at scale.

Many people know K-means clustering as a powerful clustering technique but not all listeners will be as familiar with spectral clustering. In today's episode, Sybille Hess from the Data Mining group at TU Eindhoven joins us to discuss her work around spectral clustering and how its result could potentially cause a massive shift from the conventional neural networks. Listen to learn about her findings. Visit our website for additional show notes Thanks to our sponsor, Weights & Biases

Building a fair machine learning model has become a critical consideration in today's world. In this episode, we speak with Anshuman Chabra, a Ph.D. candidate in Computer Networks. Chhabra joins us to discuss his research on building fair machine learning models and why it is important. Find out how he modeled the problem and the result found.

In today's episode, Jason, an Assistant Professor of Statistical Science at Duke University talks about his research on K power means. K power means is a newly-developed algorithm by Jason and his team, that aims to solve the problem of local minima in classical K-means, without demanding heavy computational resources. Listen to find out the outcome of Jason's study. Thanks to our Sponsors:ClearML is an open-source MLOps solution users love to customize, helping you easily Track, Orchestrate, and Automate ML workflows at scale. https://clear.ml

In this episode, Kyle interviews Lucas Murtinho about the paper "Shallow decision treees for explainable k-means clustering" about the use of decision trees to help explain the clustering partitions.  Check out our website for extended show notes and images! Thanks to our Sponsors:ClearML is an open-source MLOps solution users love to customize, helping you easily Track, Orchestrate, and Automate ML workflows at scale.

In this episode of Awake All Hours, we begin by remembering the immense legacy of Jamal Edwards before sharing our experience at Timedance's incredible 7th Birthday bash at the Trinity Centre, with three of the finest DJs on planet earth - Batu, Helena Hauff and k means!We then discuss the heightened crossover between underground electronic music and high fashion in recent months before diving into album reviews, which features yeule's deeply personal and experimental 'Glitch Princess', Slikback's punishing new collaborative project 'CONDENSE' and Silicon Scally's classy new electro LP, 'Field Lines'.We finish, as ever, with What's In My Bandcamp Basket, rounding up our favourite little morsels off the internet - club ready stuff, club-adjacent bits, compilations, represses or anything we think is well cool!SUPPORT THE ARTISTS: https://buymusic.club/list/awakeallhours-all-hail-k-means-the-crossover-between-underground-electronic-music-and-high-fashion-yeuleSUPPORT AWAKE ALL HOURS: www.patreon.com/awakeallhours

Linh Da joins us to explore how image segmentation can be done using k-means clustering.  Image segmentation involves dividing an image into a distinct set of segments.  One such approach is to do this purely on color, in which case, k-means clustering is a good option.  Thanks to our Sponsors:   Visit Weights and Biases mention Data Skeptic when you request a demo! & Nomad Data  In the image below, you can see the k-means clustering segmentation results for the same image with the values of 2, 4, 6, and 8 for k.  

In today's episode, Gregory Glatzer explained his machine learning project that involved the prediction of elephant movement and settlement, in a bid to limit the activities of poachers. He used two machine learning algorithms, DBSCAN and K-Means clustering at different stages of the project. Listen to learn about why these two techniques were useful and what conclusions could be drawn.

Welcome to our new season, Data Skeptic: k-means clustering.  Each week will feature an interview or discussion related to this classic algorithm, it's use cases, and analysis. This episode is an overview of the topic presented in several segments.

In this episode I talk about the problems of dimensionality reduction and clustering. I explain the applications of each one of these problems and also the most famous methods for solving them, such as PCA, KPCA, ICA and NNMF for the dimensionality reduction and the Kmeans for the clustering problems. In the end I also explain the autoencoders, which are powerfull neural networks that can be used for both problems. Instagram: https://www.instagram.com/podcast.lifewithai/ Linkedin: https://www.linkedin.com/company/life-with-ai Code: https://github.com/filipelauar/projects/tree/main/dimensionality%20reduction%20and%20clustering

Nesse episódio eu falo sobre os problemas de redução de dimensionalidade e clustering. Eu explico aplicações de de cada um desses problemas e também os algoritmos mais famosos pra cada um deles, como PCA, KPCA, ICA e NNMF para redução de dimensionalidade e o Kmeans para problemas de clustering. No final eu também explico os Autoencoders, que é uma arquitetura de rede neural muito poderosa que funciona para os dois problemas. Instagram: https://www.instagram.com/podcast.lifewithai/ Linkedin: https://www.linkedin.com/company/life-with-ai Códigos: https://github.com/filipelauar/projects/tree/main/dimensionality%20reduction%20and%20clustering

As we stroll into the summer, we're happy to provide you with a fresh splash by Avant Garbage. The London-based duo k means and DJ ojo got entangled over their love for peculiar sounds and offers you a monthly refill on Netil Radio , with the stranger margins of music, deftly balancing feelings of brightness and darkness. This two-hour (!) mix channels their love for disorientating dub and crunchy rhythms, combining old favourites and new digs ~ all infused with the softness of a psychedelic teabag. Rejuvenate and replenish with this one! Stay hydrated, xr @k_means @ojodj

Unsupervised learning for finding fraud cases --- Send in a voice message: https://anchor.fm/david-nishimoto/message

A discussion of the role of analytics in segmentation, targeting, and positioning.

Michelle Martin speaks to Arun Pai, Chief Strategy Officer at Flow to discuss Bitcoin's long-term prospects and Goldman and Citi's projection for Gold in 2021.  See omnystudio.com/listener for privacy information.

Link to bioRxiv paper: http://biorxiv.org/cgi/content/short/2020.11.15.383281v1?rss=1 Authors: Oh, S., Park, H., Zhang, X. Abstract: Recent advances in single cell transcriptomics have allowed us to examine the identify of each single cell, thus have led to discovery of new cell types and provide a high resolution map of cell type composition in tissues. Technologies which can measure another type of data of a single cell in addition to the gene-expression data provide a more comprehensive picture of a cell, and meanwhile pose challenges for data integration tasks. We consider the spatial location of cells, which is an important feature of cells, combined with the cells' gene-expression profiles, to determine the cell type identity. We aim to jointly classify cells based on their locations relative to other cells in the system as well as their gene expression profiles. We have developed scHybridNMF (single-cell Hybrid Nonnegative Matrix Factorization), which performs cell type identification by incorporating single cell gene expression data with cell location data. We combined two classical methods, nonnegative matrix factorization with a k-means clustering scheme, to respectively represent high-dimensional gene expression data and low-dimensional location data together. Our method incorporates a novel cell location term to the gene expression clustering. We show that scHybridNMF can make use of the location data to improve cell type clustering. In particular, we show that under multiple scenarios, including that when the number of genes profiled is low, and when the location data is noisy, scHybridNMF outperforms the standalone algorithms NMF and k-means, and an existing method HMRF which also uses cell location and gene-expression data for cell type identification. Copy rights belong to original authors. Visit the link for more info

Kmeans (sklearn vs FAISS), finding n_clusters via inertia/silhouette, Agglomorative, DBSCAN/HDBSCAN

K Means has been bending the rules of UK bass music for the past few years, via her broadcasts for Noods Radio and Netil Radio, and residencies for Bristol's Psychotherapy Sessions and her own Expansion Pak club nite. Her sets alongside the likes of Anz and Quest?onmarc have sent parties across the UK into an absolute frenzy, with a signature mix of headspinning bass, slapping club edits, wild footwork and mutant dancehall. She gets supernatural for an hour. Vic Bang - Rim [Kit Records] Flatpack & Dudeman - On The Tools [Pressure Dome] Sputnik One - Shader [First Second Label] N/OBE - Cistern [Flex] Endless Mow - Fey Dor [All Centre] Yushh - Team Boot [Woozy] MoMA Ready - Deep Enough [Self] Superabundance - Antimatter Circus [Future Times] DJ Slugo - I’m Higher Than A M.F. [WIDE] Nikki Nair - One And Only [Shubzin] Filter Dread - Beyond Saturn [TV Showw] East Man & Walton - Gunshot [Hi Tek Sounds] Ice_Eyes - Dirt [ANBA] Pinch - All Man Got (ft Trim) [Tectonic] Cartridge - Trippy Mane [CNCPT Collective] Cimm - Endless Sky [Sentry Records] DJ Girl - Psychosis [Worst Behaviour Recs] Traxman - WAP Juked Out [Self] Cesrv - WATCH YO SLANG [Beatwise] Kush Jones - SYSTEMS [Self] DJ Manny - Fluorescent [Self] DJ ojo - ??? [??] Bastiengoat - Slurpee [Worst Behaviour Recs]

Do you want to learn the how and when of implementing K-means clustering in Python? Would you like to practice your pandas skills with a real-world project? This week on the show, David Amos is back with another batch of PyCoder’s Weekly articles and projects.

Link to bioRxiv paper: http://biorxiv.org/cgi/content/short/2020.07.31.231662v1?rss=1 Authors: Hu, Y., Zhang, Z. Abstract: Grain characteristics, including kernel length, kernel width, and thousand kernel weight, are critical component traits for grain yield. Manual measurements and counting are expensive, forming the bottleneck for dissecting the genetic architecture of these traits toward ultimate yield improvement. High-throughput phenotyping methods have been developed by analyzing images of kernels. However, segmenting kernels from the image background and noise artifacts or from other kernels positioned in close proximity remain challenges. In this study, we developed a software package, named GridFree, to overcome these challenges. GridFree uses an unsupervised machine learning approach, K-Means, to segment kernels from the background by using principal component analysis on both raw image channels and their color indices. GridFree incorporates users' experiences as a dynamic criterion to set thresholds for a divide-and-combine strategy that effectively segments adjacent kernels. When adjacent multiple kernels are incorrectly segmented as a single object, they form an outlier on the distribution plot of kernel area, length, and width. GridFree uses the dynamic threshold settings for splitting and merging. In addition to counting, GridFree measures kernel length, width, and area with the option of scaling with a reference object. Evaluations against existing software programs demonstrated that GridFree had the smallest error on counting seeds for multiple crops, including alfalfa, canola, lentil, wheat, chickpea, and soybean. GridFree was implemented in Python with a friendly graphical user interface to allow users easily visualize the outcomes and make decisions, which ultimately eliminates time-consuming and repetitive manual labor. GridFree is freely available at the GridFree website (https://zzlab.net/GridFree). Copy rights belong to original authors. Visit the link for more info

I talk about this weeks work completed and what I am learning in data science

Normalizing data to reduce variance is necessary preparation of data. Normalizing is rescaling they data to a standard deviation of 1. Sentiment analysis can analyze text content for negative or positive words.

Unsupervised learning

Kmeans to group similar groups of data

I explain what is an unsupervised classification algorithm and what is a supervised algorithm.  I use examples about remote sensing image classification and I discuss my opinion about the unsupervised algorithms, which are in fact similar to the supervised ones.  Take as one example the well known unsupervised K-Means algorithm. The analyst must inform a priori the most important parameter to run the algorithm, the K value. I have a video about the K-Means algorithm. Follow my podcast: http://anchor.fm/tkorting Subscribe to my YouTube channel: http://youtube.com/tkorting The intro and the final sounds were recorded at my home, using an old clock that belonged to my grandmother. Thanks for listening

An airhacks.fm conversation with Pavel Pscheidl (@PavelPscheidl) about: Pentium 1 with 12, 75 MHz, first hello world with 17, Quake 3 friend as programming coach, starting with Java 1.6 at at the university of Hradec Kralove, second "hello world" with Operation Flashpoint, the third "hello world" was a Swing Java application as introduction to object oriented programming, introduction to enterprise Java in the 3rd year at the university, first commercial banking Java EE 6 / WebLogic project in Prague with mobile devices, working full time during the study, the first Java EE project was really successful, 2 month development time, one DTO, nor superfluous layers, using enunciate to generate the REST API, CDI and JAX-RS are a strong foundation, the first beep, fast JSF, CDI and JAX-RS deployments, the first beep, the War of Frameworks, pragmatic Java EE, "no frameworks" project at telco, reverse engineering Java EE, getting questions answered at airhacks.tv, working on PhD and statistics, starting at h2o.ai, h2o is a sillicon valley startup, h2o started as a distributed key-value store with involvement of Cliff Click, machine learning algorithms were introduced on top of distributed cache - the advent of h2o, h2o is an opensource company - see github, Driverless AI is the commercial product, Driverless AI automates cumbersome tasks, all AI heavy lifting is written in Java, h2o provides a custom java.util.Map implementation as distributed cache, random forest is great for outlier detection, the computer vision library openCV, Gradient Boosting Machine (GBM), the opensource airlines dataset, monitoring Java EE request processing queues with GBM, Generalized Linear Model (GLM), GBM vs. GLM, GBM is more explained with the decision tree as output, XGBoost, at h2o XGBoost is written in C and comes with JNI Java interface, XGBoost works well on GPUs, XGBoost is like GBM but optimized for GPUs, Word2vec, Deep Learning (Neural Networks), h2o generates a directly usable archive with the trained model -- and is directly usable in Java, K-Means, k-means will try to find the answer without a teacher, AI is just predictive statistics on steroids, Isolation Random Forest, IRF was designed for outlier detection, and K-Means was not, Naïve Bayes Classifier is rarely used in practice - it assumes no relation between the features, Stacking is the combination of algorithms to improve the results, AutoML: Automatic Machine Learning, AutomML will try to find the right combination of algorithms to match the outcome, h2o provides a set of connectors: csv, JDBC, amazon S3, Google Cloud Storage, applying AI to Java EE logs, the amount of training data depends on the amount of features, for each feature you will need approx. 30 observations, h2o world - the conference, cancer prediction with machine learning, preserving wildlife with AI, using AI for spider categorization Pavel Pscheidl on twitter: @PavelPscheidl, Pavel's blog: pavel.cool

Union Home Minister Amit Shah on Monday announced the scrapping of Article 370 of the Constitution, which provides a special status to the state of Jammu and Kashmir. The government also moved a Bill proposing the bifurcation of the state into two Union Territories - first, Jammu and Kashmir, which will have a legislature like Delhi, and second Ladakh, which will not have a legislature like Chandigarh. The official notification on the matter was signed by President Ram Nath Kovind. Meanwhile, J&K remains on edge, with some of its leaders including ex-chief ministers Omar Abdullah and Mehbooba Mufti under house arrest. Listen to this podcast for more.

פודקאסט מספר 370 של רברס עם פלטפורמה - אורי ורן מארחים בכרכור את אתי גבירץ מחברת ThetaRay לשיחה על תוכנית הלימודים למכונות וילדים מוצלחים.אתי היא VP Product Management ב-ThetaRay - חברת בינה מלאכותית שמשלבת טכנולגיות Big Data עם אלגוריתמים ייחודים “שלומדים אינטואיטיבית” (Unsupervised Learning) שפותחו בחברה.הפלטפורמה משמשת אירגונים פיננסיים גלובאליים במלחמה בפשעים כלכליים (הלבנת הון, מימון טרור, סחר באנשים ושאר רעות חולות).הפתרונות גנריים לחלוטין ויכולים לשמש גם למקרים אחרים, אבל כרגע המיקוד הוא בתחום הפיננסי.שתי שאלות כלליות לפני הצלילה לטכנולוגיה - הלבנת הון נשמע אכן קשור לנושאים פיננסיים - איך סחר בנשים (למשל) מתקשר?בסופו של דבר צריך להעביר את הכסף . . .לחברה יש מיזם Pro bono עם עמותה בארה”ב, שמחפשת סימנים לסחר באנשים באמצעות מאגר מידע גדול, שחלקו כלכלי, ו-ThetaRay עוזרת למצוא נקודות שצריך לבדוק.לא מעט פשעים לאחרונה מתבצעים תוך שימוש במטבעות קריפטוגרפיים (נכון?) - האם יש ממשק גם לכיוון הזה?התחושה כנראה נכונה, אין כרגע ממשק פעיל בתחום אבל בהחלט יש מחקרים.אז נתחיל - מה זה Machine Learning מבחינתכם? איך זה משרת את החברה?ראשית - מוטיבציה: למה מכונות צריכות בכלל ללמוד?דמיינו ערימה של חפצים בצבעים שונים. למען הפשטות - רק צבעים אחידים, ואף אחד לא עיוור-צבעים (אין גברים בקהל, נכון?)למיין חמישים פריטים כאלה בשעה - לא בעיה, ונשאר המון זמןלמיין “הר” של כאלה (כמה מיליארדים) - כן בעיה, לא בשעה ולא ביום, חוץ מזה שגם ממש לא בא לכם לעשות את זה אלא לתת למישהו אחר (משימה פשוטה וחזרתית עם צורך ב-Throughput גבוה - מזכיר את פרק 363 על ה-GPU).אנחנו כבר יודעים למיין - “רק” צריך ללמד את המכונה לעשות את זה. איך? כמו שמלמדים ילדים: “זה תפוח אדום”, “זה אגס צהוב” וכו’. בפעם הבאה שואלים “מה זה?” ונותנים פידבק על התשובה, עד שיוסי הילד המוצלח לומד באמצעות דוגמאות - גם על המקרה הספציפי וגם להשליך על דברים אחרים (“כרבולת של תרנגול זה גם אדום” וכו’).כולל לרוב גם מקרים של “תרנגול זה תפוח” ודיונים על טבעונות, אבל זה כבר עניין אחר.באופן דומה ניתן ללמד מכונה להבדיל בין צבעים למשל - מתן המון דוגמאות ואז בחינה של התוצאה, תיקון ושוב עד לתוצאה הרצויה - ועכשיו המכונה יודעת למיין את הפריטים לפי צבע במקומכם.השלב הבא: באמצע הערימה יש פריט סגול . . כזה לא היה לנו קודם. מה עכשיו?אפשר להגיד “זה דומה לכחול” ואז לסווג ככחול בסבירות בינונית. האם זו טעות? תלוי בהגדרה.אפשר להגיד “זה דומה לכחול וגם לאדום” ולהגיד שלא ניתן לסווג בבטחון. האם זו טעות? שוב. תלוי מה רוצים, ומה הוא רווח הסמך שהוגדר.המקביל בדוגמא שלנו - יוסי לא יודע מה זה חציל כי לא היה קודם (ועוד לא ביקר אצל עובד).מה שראינו כאן זה דוגמא למגבלה: המכונה “לא יודעת” את מה שלא “לימדו” אותה קודם - Supervised Learning: מישהו מפקח על הלמידה.יש הגדרה של ה-Training Set - הקטיגוריות שיכללו, כמה דוגמאות בכל אחת, מהם ה - thresholds ולזיהוי כו’.המדדים להצלחה הם דיוק (Accuracy, Precision) וכיסוי (Coverage, Recall).מבחינת Detection יש התייחסות גם ל Detection Rateהכל חשוב ברמה העסקית - מהי המטרה שהמכונה משרתת? כאן יש גם כלים כמו ROC Curve או Confusion matrix שבאמצעותם מגדירים את הסף הנדרשמה חשוב כאן יותר - דיוק או כיסוי?מכונה שלומדת “טוב יותר” מצריכה פחות פשרות - אבל תמיד יש דעיכה (deterioration) וככל שמתרחקים מקבוצת הלמידה יש סיכוי גבוה יותר לטעות.חוץ מאצל יוסי.וכמו את יוסי - גם את המכונה צריך להמשיך ללמד.אז מה לגבי Unsupervised Learning?כל מה שדיברנו עליו עד עכשיו מוגדר כ”למידה קונספטואלית”. על מנת להבין Unsupervised Learning, נלך שוב לדוגמא של לימוד ילדים - והפעם באמצעות התבוננות (Observation) ולמידה אינטואיטיבית, בלי שמישהו אחר יגדיר זאת עבורם מראש.אחד הדברים שנלמדים כך זה ההגדרה של “נורמטיבי” - ומהי אנומליה.דוגמא - דני בן ה-3, כבר “ראה עולם” (בכל זאת, בן 3) וגם כבר יודע להתנסח ולהסיק מסקנות. יום אחד הוא רואה בסופרמרקט בפעם הראשונה אדם בכסא גלגלים - ומצביע - “שונה”. למה שונה? “כי הוא הולך עם גלגלים במקום רגליים”.יש כאן שני דברים - (1) זיהוי של אנומליה ו (2) הסבר (Evidence, Forensics) - ב-ThetaRay זה מכונה Trigger Feature.לפי מה הילד החליט? ע”פ הנורמה ולפי מה שהוא נחשף אליו קודם (הידוע לו עד כה).המשך הדוגמא - באותו הסופרמרקט נמצאת גם גלית, שגרה באותה השכונה - ובמשפחה שלה יש דוד שנעזר בכסא גלגלים. גלית עוברת באותו מקום ולא מצביעה.כתלות בהגדרת המשימה - דני דייק (וזיהה) וגלית פיספסה (בהנחה שהמשימה היא למצוא אנומליה ולא לסווג לקבוצות חדשות). לגלית לא “הפריע” שום דבר.חשוב להבחין כאן בין Unsupervised ל-Ungoverned- גם גלית וגם דני נחשפו לעולם ע”י ההורים שלהם, והלמידה שלהם הייתה “ממושטרת” (Governed) - ההורים החליטו על חוגים, על טיולים, מקום מגורים, צפייה בטלוויזייה וכו’. גלית גדלה במשפחה שהשפיעה על יכולת שלה לזהות (למשל) ששימוש בכסא גלגלים הוא משהו שקורה באחוזים קטנים יחסית באוכלוסיה (שוב - תלוי בהגדרות של איזו אוכלוסיה ומתי ואיך).בתנאים הללו ותחת ההגדרות הללו - דני יצליח יותר מגלית (יש קשר בין האופן בו האלגוריתם לומד והגדרת המשימה).ברמה הפילוסופית יש כאן משהו מעניין - דני הצליח כי הוא לא ראה דוגמא כזו בעבר, וחגית נכשלה כי היא כן ראתה דוגמא כזו - שזה הפוך מהדוגמא הקודמת (יוסי נכשל כי לא ראה “סגול” לפני כן).בחזרה לענייני ההגדרות - אם ניקח אלגוריתם שמטרתו לקחת ערימה של נתונים (הצעות למכירת מוצרים למשל) ולסווג אותם (Clustering) - אופן הביצוע דומה: בכל פעם שנתקלים במשהו חדש יוצרים קבוצה חדשה.זו עוד דוגמא ל Unsupervised Learning, שנקראת Clustering  - היכולת למצוא באוכלוסיה קטיגוריות שאינן מוגדרות מראש (שזה באיזשהו מקום קצת “לחפש מתחת לפנס”).אם אנחנו יודעים להגדיר מהי “אוכלוסיה מייצגת” (שזו אומנות בפני עצמה - ד”ש לגילדת ה-Data Scientists), אפשר לקחת אלגוריתם Clustering (שהוא Unsupervised), שלא יודע כמה קטיגוריות יש בתוך האוכלוסיה (אז פסלנו את K-Means, שמניח מספר קטיגוריות מוגדר), ולהריץ על מנת “למצוא קבוצות דומות”, ע”פ ספים מוגדרים של שונות, או ע”י הגדרת סוג המימדים שמעניינים אותנו.דוגמא: ב-ThetaRay מזהים אנומליות - ומזהים פשעים, מתוך הנחה שהאזרחים באופן כללי שומרי חוק, ותוך הנחה שפשע הוא אנומליהזה לא תמיד נכון, ולכל מדינה יש את הספים שלה - דוגמא שבה לקוח פוטנציאלי אמר שיש לו מערכות במדינה כלשהי שבה 40% מהאוכלוסיה מעורבות בהונאות (בדרום אמריקה! לא כאן, מה פתאום?). יש כאן הפרה של הנחת הבסיס והמערכת כנראה לא  תיהיה יעילה (כי זה גם לא רוב מוחלט שמאפשר לזהות את הקבוצה המשלימה).מרגע שזיהינו אנומליות, השלב הבא הוא להפוך אותן למשהו שאפשר לעבוד איתו (Actionable). על מנת שאנליסט יוכל לחקור הלבנת הון למשל, צריך לתאר לו (פורמאלית) מה זה.אלו ה - Trigger Features שהזכרנו קודם - וזה משהו שצריך להיות מסוגלים להסביר (בסופו של דבר אלו תיאורים מתימטיים שהוגדרו בדר”כ ע”י Data Scientist)יש כאן אלמנט של Feature Engineering - התהליך בסופו של דבר מנוהל (Governed), וה - Data Scientist משמש כ”הורה” בעולם בו האלגוריתמים לומדים (ראיתם Person of Interest? אז Finch).התחלנו עם למה בכלל צריך Machine Learning, המשכנו להבדל בין למידה שהיא Supervised לעומת Unsupervised ואז ספציפית לתוך Governed Unsupervised Learning ע”י הכתבה של Features.האם מדובר בהגדרות ספציפיות (“זה הטקטסט שאנחנו מחפשים”) או בהגדרות כלליות של שפה וכלים?בפתרון אין heuristics או הגדרות של מה “נורמאלי” - ה-Features מוגדרים כמימדים בעולם, והאלגוריתם לומד אותם.המטרה היא לזהות פשעים ולא חצילים - צריך להנגיש את המידע הרלוונטי באופן שיאפשר לזהות אנומליות באופן מיטבי (ביחס לאובייקטים אחרים וביחס לעצמו).צריך גם להגדיר מהי טרנזאקציה כספית, מה האפשרויות וכל מידע אחר שאפשר להנגיש (מי האדם? מי השותפים האפשריים שלו? ועוד)בשלב הבא - Clustering ומציאת דימיון בין האנומליות ומציאת צורות התנהגות חדשות שלא חשבנו עליהן מראש.הסיווג הוא לשתי קבוצות

Deze week bespreken de millenials het leven. Wijn van de week: Marques de Cacereshttps://www.flesjewijn.com/wijnen/marques-de-caceres-crianza-325 Een korte geschiedenis van de tijd - Stephen Hawking: https://www.bibliotheek.nl/catalogus/titel.39759433X.html/een-korte-geschiedenis-van-de-tijd/ De large kledingwinkel: https://www.large.nl/ De subreddit van Fontys Hogeschool ICT: https://www.reddit.com/r/FHICT/ Onze homescreens: https://imgur.com/gallery/rsT5FZQ Film van volgende week: V for Vendetta

We speak with David Kopec, professor & iOS developer consultant, about his book "Classic Computer Science Problems in Swift". What can you learn from solving classic CS problems such as search, constraint-satisfaction, graph problems and more? David gives a brief explanation of some of the interesting problems in the book such as K-Means clustering and Genetic algorithms. Use promo code 'pckopec' at https://www.manning.com/books/classic-computer-science-problems-in-swift to purchase the book for half price! Wanna chat with other smart iOS developers? Sign up for our free forum: https://forum.insideiosdev.com

This 70th episode of Learning Machines 101 we discuss how to identify facial emotion expressions in images using an advanced clustering technique called Stochastic Neighborhood Embedding. We discuss the concept of recognizing facial emotions in images including applications to problems such as: improving online communication quality, identifying suspicious individuals such as terrorists using video cameras, improving lie detector tests, improving athletic performance by providing emotion feedback, and designing smart advertising which can look at the customer’s face to determine if they are bored or interested and dynamically adapt the advertising accordingly. To address this problem we review clustering algorithm methods including K-means clustering, Linear Discriminant Analysis, Spectral Clustering, and the relatively new technique of Stochastic Neighborhood Embedding (SNE) clustering. At the end of this podcast we provide a brief review of the classic machine learning text by Christopher Bishop titled “Pattern Recognition and Machine Learning”. Make sure to visit: www.learningmachines101.com to obtain free transcripts of this podcast and important supplemental reference materials!

Looking to maximize the effectiveness of your data analytical models? From RFM segmentation to K-Means segmentation, learn how asset managers are sourcing data and measuring success.

This recitation reviews hierarchical and kmeans clustering in great detail with many different code examples. It also discusses the merits and disadvantages of both clustering and how to calculate error.

Evaluation of a new k-means approach for exploratory clustering of items

The k-means clustering algorithm is an algorithm that computes a deterministic label for a given "k" number of clusters from an n-dimensional datset.  This mini-episode explores how Yoshi, our lilac crowned amazon's biological processes might be a useful way of measuring where she sits when there are no humans around.  Listen to find out how!