andrew_gelman_stats andrew_gelman_stats-2013 andrew_gelman_stats-2013-1740 knowledge-graph by maker-knowledge-mining
Source: html
Introduction: Following up on our previous post , Andrew Wilson writes: I agree we are in a really exciting time for statistics and machine learning. There has been a lot of talk lately comparing machine learning with statistics. I am curious whether you think there are many fundamental differences between the fields, or just superficial differences — different popular approximate inference methods, slightly different popular application areas, etc. Is machine learning a subset of statistics? In the paper we discuss how we think machine learning is fundamentally about pattern discovery, and ultimately, fully automating the learning and decision making process. In other words, whatever a human does when he or she uses tools to analyze data, can be written down algorithmically and automated on a computer. I am not sure if the ambitions are similar in statistics — and I don’t have any conventional statistics background, which makes it harder to tell. I think it’s an interesting discussion.
sentIndex sentText sentNum sentScore
1 Following up on our previous post , Andrew Wilson writes: I agree we are in a really exciting time for statistics and machine learning. [sent-1, score-0.534]
2 There has been a lot of talk lately comparing machine learning with statistics. [sent-2, score-0.815]
3 I am curious whether you think there are many fundamental differences between the fields, or just superficial differences — different popular approximate inference methods, slightly different popular application areas, etc. [sent-3, score-0.648]
4 In the paper we discuss how we think machine learning is fundamentally about pattern discovery, and ultimately, fully automating the learning and decision making process. [sent-5, score-1.641]
5 In other words, whatever a human does when he or she uses tools to analyze data, can be written down algorithmically and automated on a computer. [sent-6, score-0.208]
6 I am not sure if the ambitions are similar in statistics — and I don’t have any conventional statistics background, which makes it harder to tell. [sent-7, score-0.408]
7 My reply: I don’t know enough about machine learning to know what differences there are between the fields. [sent-9, score-0.797]
8 One of my sayings is that theoretical statistics is another name for the theory of applied statistics. [sent-10, score-0.468]
9 That is, statistics is all about modeling what we do, and modeling what we should be doing. [sent-11, score-0.47]
10 As always in the social sciences, normative modeling has a descriptive flavor and descriptive modeling has a normative flavor: to the extent that we’re not doing what we say we should be doing, this suggests potential changes in our theory or in our practice. [sent-12, score-1.411]
11 And much of my work over the years has been to give theoretical foundations for various areas of statistical practice that have typically been treated informally. [sent-13, score-0.279]
12 Thus, compared to other academic statisticians, I think I spend more time monitoring convergence of my iterative simulations, checking the fit of my models, and graphing data and fitted curves—but at the same time I do these things more formally than many statisticians have been trained to do. [sent-14, score-0.805]
13 I think that some of the research we’ve been discussing lately on automatic model construction (done by people other than me, let me emphasize! [sent-15, score-0.459]
14 ) is important in that is moving toward a better description—and thus also a better normative theory—of model building. [sent-16, score-0.405]
15 To me, it’s a big step forward from that thing where “learning a model” is associated with taking a big multivariate dataset and trying to identify conditional independence structures. [sent-17, score-0.258]
16 To me, all that stuff is static, and I’m much happier with a framework in which models are built out recursively in a language-like fashion. [sent-18, score-0.343]
17 We still have a ways to go in fitting models that we’ve already specified. [sent-20, score-0.087]
18 Are we at the stage of “fully automating the learning and decision making process”? [sent-22, score-0.792]
19 But the only way forward is to try, not getting too stuck in our current understanding at any time. [sent-24, score-0.108]
wordName wordTfidf (topN-words)
[('learning', 0.346), ('machine', 0.312), ('normative', 0.257), ('automating', 0.203), ('flavor', 0.171), ('modeling', 0.162), ('lately', 0.157), ('statistics', 0.146), ('descriptive', 0.142), ('differences', 0.139), ('theory', 0.118), ('algorithmically', 0.116), ('ambitions', 0.116), ('fully', 0.115), ('recursively', 0.11), ('forward', 0.108), ('sayings', 0.105), ('areas', 0.104), ('decision', 0.1), ('theoretical', 0.099), ('superficial', 0.098), ('popular', 0.098), ('statisticians', 0.095), ('automated', 0.092), ('static', 0.09), ('models', 0.087), ('monitoring', 0.087), ('wilson', 0.084), ('graphing', 0.082), ('iterative', 0.082), ('independence', 0.081), ('curves', 0.081), ('automatic', 0.079), ('thus', 0.079), ('happier', 0.078), ('construction', 0.078), ('convergence', 0.078), ('formally', 0.077), ('trained', 0.076), ('time', 0.076), ('think', 0.076), ('foundations', 0.076), ('fundamentally', 0.073), ('stage', 0.073), ('subset', 0.072), ('making', 0.07), ('multivariate', 0.069), ('model', 0.069), ('discovery', 0.068), ('built', 0.068)]
simIndex simValue blogId blogTitle
same-blog 1 1.0000001 1740 andrew gelman stats-2013-02-26-“Is machine learning a subset of statistics?”
Introduction: Following up on our previous post , Andrew Wilson writes: I agree we are in a really exciting time for statistics and machine learning. There has been a lot of talk lately comparing machine learning with statistics. I am curious whether you think there are many fundamental differences between the fields, or just superficial differences — different popular approximate inference methods, slightly different popular application areas, etc. Is machine learning a subset of statistics? In the paper we discuss how we think machine learning is fundamentally about pattern discovery, and ultimately, fully automating the learning and decision making process. In other words, whatever a human does when he or she uses tools to analyze data, can be written down algorithmically and automated on a computer. I am not sure if the ambitions are similar in statistics — and I don’t have any conventional statistics background, which makes it harder to tell. I think it’s an interesting discussion.
2 0.22344738 1131 andrew gelman stats-2012-01-20-Stan: A (Bayesian) Directed Graphical Model Compiler
Introduction: Here’s Bob’s talk from the NYC machine learning meetup . And here’s Stan himself:
3 0.21062808 1482 andrew gelman stats-2012-09-04-Model checking and model understanding in machine learning
Introduction: Last month I wrote : Computer scientists are often brilliant but they can be unfamiliar with what is done in the worlds of data collection and analysis. This goes the other way too: statisticians such as myself can look pretty awkward, reinventing (or failing to reinvent) various wheels when we write computer programs or, even worse, try to design software.Andrew MacNamara writes: Andrew MacNamara followed up with some thoughts: I [MacNamara] had some basic statistics training through my MBA program, after having completed an undergrad degree in computer science. Since then I’ve been very interested in learning more about statistical techniques, including things like GLM and censored data analyses as well as machine learning topics like neural nets, SVMs, etc. I began following your blog after some research into Bayesian analysis topics and I am trying to dig deeper on that side of things. One thing I have noticed is that there seems to be a distinction between data analysi
Introduction: David Duvenaud writes: I’ve been following your recent discussions about how an AI could do statistics [see also here ]. I was especially excited about your suggestion for new statistical methods using “a language-like approach to recursively creating new models from a specified list of distributions and transformations, and an automatic approach to checking model fit.” Your discussion of these ideas was exciting to me and my colleagues because we recently did some work taking a step in this direction, automatically searching through a grammar over Gaussian process regression models. Roger Grosse previously did the same thing , but over matrix decomposition models using held-out predictive likelihood to check model fit. These are both examples of automatic Bayesian model-building by a search over more and more complex models, as you suggested. One nice thing is that both grammars include lots of standard models for free, and they seem to work pretty well, although the
5 0.1747674 1297 andrew gelman stats-2012-05-03-New New York data research organizations
Introduction: In a single day, New York City obtained two data analysis/statistics/machine learning organizations: Microsoft Research New York City with John Langford (machine learning), Duncan Watts (networks), and Dave Pennock (algorithmic economics). eBay technology center focusing on data – led by Chris Dixon , the co-founder of the recommendation engine company Hunch, which has recently been acquired by eBay. New York already has Facebook’s engineering unit , Twitter’s East Coast headquarters , and Google’s second-largest engineering office. The data community here is on an upswing, and it might be one of the best places to be if you’re into applied statistics, machine learning or data analysis. Post by Aleks Jakulin . P.S. (from Andrew): The formerly-Yahoo-now-Microsoft researchers have a more-or-less formal connection to Columbia, through the Applied Statistics Center, where some of them will be organizing occasional mini-conferences and workshops!
6 0.16689508 2129 andrew gelman stats-2013-12-10-Cross-validation and Bayesian estimation of tuning parameters
7 0.16473351 1630 andrew gelman stats-2012-12-18-Postdoc positions at Microsoft Research – NYC
8 0.16271034 1992 andrew gelman stats-2013-08-21-Workshop for Women in Machine Learning
10 0.15372163 1902 andrew gelman stats-2013-06-17-Job opening at new “big data” consulting firm!
11 0.15254402 1126 andrew gelman stats-2012-01-18-Bob on Stan
12 0.14992581 2072 andrew gelman stats-2013-10-21-The future (and past) of statistical sciences
13 0.143474 2366 andrew gelman stats-2014-06-09-On deck this week
14 0.14246646 1280 andrew gelman stats-2012-04-24-Non-Bayesian analysis of Bayesian agents?
15 0.13844813 781 andrew gelman stats-2011-06-28-The holes in my philosophy of Bayesian data analysis
16 0.13835245 1431 andrew gelman stats-2012-07-27-Overfitting
17 0.13712069 1110 andrew gelman stats-2012-01-10-Jobs in statistics research! In New Jersey!
18 0.12703209 1469 andrew gelman stats-2012-08-25-Ways of knowing
19 0.12522167 65 andrew gelman stats-2010-06-03-How best to learn R?
20 0.12122904 165 andrew gelman stats-2010-07-27-Nothing is Linear, Nothing is Additive: Bayesian Models for Interactions in Social Science
topicId topicWeight
[(0, 0.229), (1, 0.068), (2, -0.097), (3, 0.044), (4, -0.007), (5, 0.089), (6, -0.127), (7, -0.002), (8, 0.037), (9, 0.081), (10, -0.039), (11, 0.028), (12, -0.023), (13, 0.0), (14, -0.065), (15, -0.022), (16, -0.011), (17, -0.031), (18, 0.002), (19, -0.034), (20, 0.031), (21, -0.083), (22, -0.012), (23, 0.044), (24, -0.022), (25, 0.038), (26, -0.032), (27, -0.001), (28, -0.024), (29, -0.012), (30, 0.001), (31, 0.011), (32, -0.005), (33, -0.025), (34, 0.01), (35, -0.029), (36, 0.013), (37, 0.012), (38, -0.063), (39, -0.032), (40, -0.054), (41, -0.031), (42, -0.002), (43, 0.056), (44, 0.02), (45, 0.026), (46, 0.044), (47, 0.007), (48, 0.058), (49, -0.06)]
simIndex simValue blogId blogTitle
same-blog 1 0.96110475 1740 andrew gelman stats-2013-02-26-“Is machine learning a subset of statistics?”
Introduction: Following up on our previous post , Andrew Wilson writes: I agree we are in a really exciting time for statistics and machine learning. There has been a lot of talk lately comparing machine learning with statistics. I am curious whether you think there are many fundamental differences between the fields, or just superficial differences — different popular approximate inference methods, slightly different popular application areas, etc. Is machine learning a subset of statistics? In the paper we discuss how we think machine learning is fundamentally about pattern discovery, and ultimately, fully automating the learning and decision making process. In other words, whatever a human does when he or she uses tools to analyze data, can be written down algorithmically and automated on a computer. I am not sure if the ambitions are similar in statistics — and I don’t have any conventional statistics background, which makes it harder to tell. I think it’s an interesting discussion.
2 0.81528157 1482 andrew gelman stats-2012-09-04-Model checking and model understanding in machine learning
Introduction: Last month I wrote : Computer scientists are often brilliant but they can be unfamiliar with what is done in the worlds of data collection and analysis. This goes the other way too: statisticians such as myself can look pretty awkward, reinventing (or failing to reinvent) various wheels when we write computer programs or, even worse, try to design software.Andrew MacNamara writes: Andrew MacNamara followed up with some thoughts: I [MacNamara] had some basic statistics training through my MBA program, after having completed an undergrad degree in computer science. Since then I’ve been very interested in learning more about statistical techniques, including things like GLM and censored data analyses as well as machine learning topics like neural nets, SVMs, etc. I began following your blog after some research into Bayesian analysis topics and I am trying to dig deeper on that side of things. One thing I have noticed is that there seems to be a distinction between data analysi
Introduction: David Duvenaud writes: I’ve been following your recent discussions about how an AI could do statistics [see also here ]. I was especially excited about your suggestion for new statistical methods using “a language-like approach to recursively creating new models from a specified list of distributions and transformations, and an automatic approach to checking model fit.” Your discussion of these ideas was exciting to me and my colleagues because we recently did some work taking a step in this direction, automatically searching through a grammar over Gaussian process regression models. Roger Grosse previously did the same thing , but over matrix decomposition models using held-out predictive likelihood to check model fit. These are both examples of automatic Bayesian model-building by a search over more and more complex models, as you suggested. One nice thing is that both grammars include lots of standard models for free, and they seem to work pretty well, although the
4 0.77688146 1788 andrew gelman stats-2013-04-04-When is there “hidden structure in data” to be discovered?
Introduction: Michael Collins sent along the following announcement for a talk: Fast learning algorithms for discovering the hidden structure in data Daniel Hsu, Microsoft Research 11am, Wednesday April 10th, Interschool lab, 7th floor CEPSR, Columbia University A major challenge in machine learning is to reliably and automatically discover hidden structure in data with minimal human intervention. For instance, one may be interested in understanding the stratification of a population into subgroups, the thematic make-up of a collection of documents, or the dynamical process governing a complex time series. Many of the core statistical estimation problems for these applications are, in general, provably intractable for both computational and statistical reasons; and therefore progress is made by shifting the focus to realistic instances that rule out the intractable cases. In this talk, I’ll describe a general computational approach for correctly estimating a wide class of statistical mod
5 0.72982156 1859 andrew gelman stats-2013-05-16-How do we choose our default methods?
Introduction: I was asked to write an article for the Committee of Presidents of Statistical Societies (COPSS) 50th anniversary volume. Here it is (it’s labeled as “Chapter 1,” which isn’t right; that’s just what came out when I used the template that was supplied). The article begins as follows: The field of statistics continues to be divided into competing schools of thought. In theory one might imagine choosing the uniquely best method for each problem as it arises, but in practice we choose for ourselves (and recom- mend to others) default principles, models, and methods to be used in a wide variety of settings. This article briefly considers the informal criteria we use to decide what methods to use and what principles to apply in statistics problems. And then I follow up with these sections: Statistics: the science of defaults Ways of knowing The pluralist’s dilemma And here’s the concluding paragraph: Statistics is a young science in which progress is being made in many
6 0.72786587 903 andrew gelman stats-2011-09-13-Duke postdoctoral fellowships in nonparametric Bayes & high-dimensional data
8 0.72757584 1282 andrew gelman stats-2012-04-26-Bad news about (some) statisticians
9 0.72401851 2151 andrew gelman stats-2013-12-27-Should statistics have a Nobel prize?
10 0.71608496 2072 andrew gelman stats-2013-10-21-The future (and past) of statistical sciences
12 0.71148926 496 andrew gelman stats-2011-01-01-Tukey’s philosophy
13 0.71060663 214 andrew gelman stats-2010-08-17-Probability-processing hardware
14 0.70926106 1076 andrew gelman stats-2011-12-21-Derman, Rodrik and the nature of statistical models
15 0.70607048 155 andrew gelman stats-2010-07-19-David Blackwell
16 0.68940276 1956 andrew gelman stats-2013-07-25-What should be in a machine learning course?
17 0.68463457 1297 andrew gelman stats-2012-05-03-New New York data research organizations
18 0.68340689 421 andrew gelman stats-2010-11-19-Just chaid
19 0.68323421 1718 andrew gelman stats-2013-02-11-Toward a framework for automatic model building
20 0.68304324 1270 andrew gelman stats-2012-04-19-Demystifying Blup
topicId topicWeight
[(1, 0.06), (5, 0.025), (9, 0.016), (15, 0.014), (16, 0.056), (21, 0.011), (24, 0.131), (66, 0.024), (73, 0.016), (84, 0.02), (86, 0.068), (95, 0.018), (99, 0.438)]
simIndex simValue blogId blogTitle
same-blog 1 0.99042571 1740 andrew gelman stats-2013-02-26-“Is machine learning a subset of statistics?”
Introduction: Following up on our previous post , Andrew Wilson writes: I agree we are in a really exciting time for statistics and machine learning. There has been a lot of talk lately comparing machine learning with statistics. I am curious whether you think there are many fundamental differences between the fields, or just superficial differences — different popular approximate inference methods, slightly different popular application areas, etc. Is machine learning a subset of statistics? In the paper we discuss how we think machine learning is fundamentally about pattern discovery, and ultimately, fully automating the learning and decision making process. In other words, whatever a human does when he or she uses tools to analyze data, can be written down algorithmically and automated on a computer. I am not sure if the ambitions are similar in statistics — and I don’t have any conventional statistics background, which makes it harder to tell. I think it’s an interesting discussion.
2 0.9883222 906 andrew gelman stats-2011-09-14-Another day, another stats postdoc
Introduction: This post is from Phil Price. I work in the Environmental Energy Technologies Division at Lawrence Berkeley National Laboratory, and I am looking for a postdoc who knows substantially more than I do about time-series modeling; in practice this probably means someone whose dissertation work involved that sort of thing. The work involves developing models to predict and/or forecast the time-dependent energy use in buildings, given historical data and some covariates such as outdoor temperature. Simple regression approaches (e.g. using time-of-week indicator variables, plus outdoor temperature) work fine for a lot of things, but we still have a variety of problems. To give one example, sometimes building behavior changes — due to retrofits, or a change in occupant behavior — so that a single model won’t fit well over a long time period. We want to recognize these changes automatically . We have many other issues besides: heteroskedasticity, need for good uncertainty estimates, abilit
3 0.98744011 738 andrew gelman stats-2011-05-30-Works well versus well understood
Introduction: John Cook discusses the John Tukey quote, “The test of a good procedure is how well it works, not how well it is understood.” Cook writes: At some level, it’s hard to argue against this. Statistical procedures operate on empirical data, so it makes sense that the procedures themselves be evaluated empirically. But I [Cook] question whether we really know that a statistical procedure works well if it isn’t well understood. Specifically, I’m skeptical of complex statistical methods whose only credentials are a handful of simulations. “We don’t have any theoretical results, buy hey, it works well in practice. Just look at the simulations.” Every method works well on the scenarios its author publishes, almost by definition. If the method didn’t handle a scenario well, the author would publish a different scenario. I agree with Cook but would give a slightly different emphasis. I’d say that a lot of methods can work when they are done well. See the second meta-principle liste
4 0.98587555 315 andrew gelman stats-2010-10-03-He doesn’t trust the fit . . . r=.999
Introduction: I received the following question from an education researcher: I was wondering if I could ask you a question about an HLM model I’m working on. The basic design is that we have 5 years of 8th grade student achievement data (standardized test scores, this is the dependent variable), 4th grade test scores, demographics (e.g., gender and ethnicity) and status wrt special ed or ELL, etc.. In addition, we have some school- or second-level information such as school averages of the student information, type of school (grade configuration), enrollment and so. In total there are thousands of students and many schools over the 5 years of information. The model we’re using is quite parsimonious, using only 7 student-level effects and 4 school-level effects. What’s puzzling us is that the correlation between predicted and actual is unrealistically high…r=0.999. We’re using the HPMIXED procedure in SAS but that shouldn’t matter. By dropping variables, obviously we can get the corre
5 0.98565209 697 andrew gelman stats-2011-05-05-A statistician rereads Bill James
Introduction: Ben Lindbergh invited me to write an article for Baseball Prospectus. I first sent him this item on the differences between baseball and politics but he said it was too political for them. I then sent him this review of a book on baseball’s greatest fielders but he said they already had someone slotted to review that book. Then I sent him some reflections on the great Bill James and he published it ! If anybody out there knows Bill James, please send this on to him: I have some questions at the end that I’m curious about. Here’s how it begins: I read my first Bill James book in 1984, took my first statistics class in 1985, and began graduate study in statistics the next year. Besides giving me the opportunity to study with the best applied statistician of the late 20th century (Don Rubin) and the best theoretical statistician of the early 21st (Xiao-Li Meng), going to graduate school at Harvard in 1986 gave me the opportunity to sit in a basement room one evening that
7 0.9844442 272 andrew gelman stats-2010-09-13-Ross Ihaka to R: Drop Dead
9 0.98228836 2009 andrew gelman stats-2013-09-05-A locally organized online BDA course on G+ hangout?
10 0.98183811 1544 andrew gelman stats-2012-10-22-Is it meaningful to talk about a probability of “65.7%” that Obama will win the election?
11 0.98164833 731 andrew gelman stats-2011-05-26-Lottery probability update
12 0.981493 1611 andrew gelman stats-2012-12-07-Feedback on my Bayesian Data Analysis class at Columbia
13 0.98117638 1323 andrew gelman stats-2012-05-16-Question 6 of my final exam for Design and Analysis of Sample Surveys
14 0.98105907 1807 andrew gelman stats-2013-04-17-Data problems, coding errors…what can be done?
15 0.98101455 1095 andrew gelman stats-2012-01-01-Martin and Liu: Probabilistic inference based on consistency of model with data
16 0.98096895 822 andrew gelman stats-2011-07-26-Any good articles on the use of error bars?
17 0.98080993 1722 andrew gelman stats-2013-02-14-Statistics for firefighters: update
18 0.98067302 1469 andrew gelman stats-2012-08-25-Ways of knowing
19 0.9803074 2258 andrew gelman stats-2014-03-21-Random matrices in the news
20 0.98009372 2041 andrew gelman stats-2013-09-27-Setting up Jitts online