Learning to Love Bayesian Statistics

At Strata NYC 2015, O'Reilly Media's data science conference, I gave a talk called "Learning to Love Bayesian Statistics".  The video is available now:


The sound quality is not great, and the clicker to advance the slides was a little wonky, but other than that, it went pretty well.

Here are the slides, if you want to follow along at home.


I borrowed illustrations from The Phantom Tollbooth, in a way I think it consistent with fair use.  I think they work remarkably well.

Thanks to the folks at Strata for inviting me to present, and for allowing me to make the video freely available.  It's actually the first video I have posted on YouTube.  I'm a late adopter.


Komentar

Posting Komentar

Postingan populer dari blog ini

Internet use and religion, part three

Gambar
This article reports preliminary results from an exploration of the relationship between religion and Internet use in Europe, using data from the European Social Survey (ESS). I describe the data processing pipeline and models in this previous article.  All the code for this article is in this IPython notebook. Data inventory The dependent variables I use in the models are rlgblg : Do you consider yourself as belonging to any particular religion or denomination? rlgdgr : Regardless of whether you belong to a particular religion, how religious would you say you are?  Scale from 0 = Not at all religious to 10 = Very religious. The explanatory variables are yrbrn : And in what year were you born? hincrank : Household income, rank from 0-1 indicating where this respondent falls relative to respondents from the same country, same round of interviews. edurank : Years of education,  rank from 0-1 indicating where this respondent falls relative to responden...
Baca selengkapnya

Recidivism and logistic regression

Gambar
In my previous article, I presented the problem of estimating a criminal's risk of recidivism, focusing on the philosophical problems of attributing a probability to an individual.  In this article I turn to the more practical problem of doing the estimation. My collaborator asked me to compare two methods and explain their assumptions, properties, pros and cons: 1) Logistic regression : This is the standard method in the field.  Researchers have designed a survey instrument that assigns each offender a score from -3 to 12.  Using data from previous offenders, they estimate the parameters of a logistic regression model and generate a predicted probability of recidivism for each score. 2) Post test probabilities : An alternative of interest to my collaborator is the idea of treating survey scores the way some tests are used in medicine, with a threshold that distinguishes positive and negative results.  Then for people who "test positive" we can compute the post...
Baca selengkapnya

One million is a lot

Gambar
When I was in third grade, the principal of my elementary school announced a bottle cap drive with the goal of collecting one million bottle caps.  The point, as I recall, was to demonstrate that one million is a very large number.  After a few months, we ran out of storage space, the drive was cancelled, and we had to settle for the lesson that 100,000 is a lot of bottle caps. So it is a special pleasure for me to announce that, early Sunday morning (November 1, 2015), this blog reached one million page views.  I am celebrating the occasion with a review of some of my favorite articles and, of course, some analysis of the page view statistics. Here's a screenshot of my Blogger stats page to make it official: And here are links to the 10 most read articles: Posts Entry Pageviews Are first babies more likely to be late? Feb 7, 2011, 9 comments 130446 All your Bayes are belong to us! Oct 27, 2011, 56 comments 47...
Baca selengkapnya