TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Statistical Modeling: The Two Cultures (2001) [pdf]

200 pointsby michael_fineabout 6 years ago

12 comments

mturmonabout 6 years ago
This is a great little paper -- with comments and rejoinder! I presented it to a reading group back when it appeared, I enjoyed it so much. Always worth re-reading because Breiman is such a hero of useful probabilistic modeling and insight.<p>One should remember that it is a reflection of its time, and the dichotomy it proposed has been softened over the years.<p>Another paper, more recent, and re-examining some of these same trends in broader context, is by David Donoho:<p><a href="https:&#x2F;&#x2F;courses.csail.mit.edu&#x2F;18.337&#x2F;2015&#x2F;docs&#x2F;50YearsDataScience.pdf" rel="nofollow">https:&#x2F;&#x2F;courses.csail.mit.edu&#x2F;18.337&#x2F;2015&#x2F;docs&#x2F;50YearsDataSc...</a><p>Highly recommended. Pretty good HN comments at:<p><a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=10431617" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=10431617</a>
评论 #19836502 未加载
ivan_ahabout 6 years ago
This is a great paper. Very long, but worth every bit of it. BTW, here is a recent blog post about the paper: <a href="http:&#x2F;&#x2F;duboue.net&#x2F;blog27.html" rel="nofollow">http:&#x2F;&#x2F;duboue.net&#x2F;blog27.html</a><p>One of the key insights I took away was the importance of using out-of-sample predictive accuracy as a metric for regression tasks in statistics—just like in ML. The standard best practices in STATS 101 is to compute R^2 coefficient (based on data of the sample), which is akin to reporting error estimates on your training data (in-sample predictive accuracy).<p>IMHO, statistics is one of the most fascinating and useful fields of study with countless applications. If only we could easily tell apart what is &quot;legacy code&quot; vs. what is fundamental... See this recent article <a href="https:&#x2F;&#x2F;www.gwern.net&#x2F;Everything" rel="nofollow">https:&#x2F;&#x2F;www.gwern.net&#x2F;Everything</a> the points out the limitations of Null Statistical Hypothesis Testing (NHST), another one of the pillars of STATS 101.
评论 #19838148 未加载
评论 #19841340 未加载
评论 #19838844 未加载
astazangastaabout 6 years ago
The title is a reference to this famous essay by C.P. Snow about a split between the humanities and science: <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;The_Two_Cultures" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;The_Two_Cultures</a>
incompatibleabout 6 years ago
See also &quot;50 years of Data Science&quot; by David Donoho (2015), which discusses the question of whether there&#x27;s any difference between &quot;statistics&quot; and &quot;data science&quot;.<p><a href="http:&#x2F;&#x2F;courses.csail.mit.edu&#x2F;18.337&#x2F;2015&#x2F;docs&#x2F;50YearsDataScience.pdf" rel="nofollow">http:&#x2F;&#x2F;courses.csail.mit.edu&#x2F;18.337&#x2F;2015&#x2F;docs&#x2F;50YearsDataSci...</a>
staredabout 6 years ago
Two? Two?!<p>There is a classic post here <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=10954508" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=10954508</a>:<p>&quot; The Geneticists: Use evolutionary principles to have a model organize itself The Bayesians: Pick good priors and use Bayesian statistics<p>The Symbolists: Use top-down approaches to modeling cognition, using symbols and hand-crafted features<p>The Conspirators: Hinton, Lecun, Bengio et al. End-to-end deep learning without manual feature engineering<p>The Swiss School: Schmidhuber et al. LSTM&#x27;s as a path to general AI.<p>The Russians: Use Support Vector Machines and its strong theoretical foundation<p>The Competitors: Only care about performance and generalization robustness. Not shy to build extremely slow and complex models.<p>The Speed Freaks: Care about fast convergence, simplicity, online learning, ease of use, scalability.<p>The Tree Huggers: Use mostly tree-based models, like Random Forests and Gradient Boosted Decision Trees<p>The Compressors: View cognition as compression. Compressed sensing, approximate matrix factorization<p>The Kitchen-sinkers: View learning as brute-force computation. Throw lots of feature transforms and random models and kernels at a problem<p>The Reinforcement learners: Look for feedback loops to add to the problem definition. The environment of the model is important.<p>The Complexities: Use methods and approaches from physics, dynamical systems and complexity&#x2F;information theory.<p>The Theorists: Will not use a method, if there is no clear theory to explain it<p>The Pragmatists: Will use an effective method, to show that there needs to be a theory to explain it<p>The Cognitive Scientists: Build machine learning models to better understand (human) cognition<p>The Doom-sayers: ML Practitioners who worry about the singularity and care about beating human performance<p>The Socialists: View machine learning as a possible danger to society. Study algorithmic bias.<p>The Engineers: Worry about implementation, pipe-line jungles, drift, data quality.<p>The Combiners: Try to use the strengths of different approaches, while eliminating their weaknesses.<p>The Pac Learners: Search for the best hypothesis that is both accurate and computationally tractable. &quot;
评论 #19840295 未加载
thousandautumnsabout 6 years ago
God I hate this paper. Perhaps it was relevant at its time. But that was 18 years ago. The described dichotomy between the &quot;two cultures&quot; isn&#x27;t nearly as pronounced, if it even exists, today. There are few statisticians today who adhere entirely to the &quot;data modeling culture&quot; as described by Breiman.<p>I&#x27;m surprised how often this paper continues to get trotted out. In my experience it seems to be a favorite of non-statisticians who use it as evidence that statistics is a dying dinosaur of a field to be superseded by X (usually machine learning). Perhaps they think if its repeated enough it will be spoken into existence?
brian_spieringabout 6 years ago
Here is a previous discussion <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=10635631" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=10635631</a>
michalrichards9about 6 years ago
I am not an expert and am still reading thru the article, but why is it such a strong dichotomy? Don&#x27;t all predictive algorithm also assume a data model? for example aren&#x27;t hidden Markov models, by assuming constant transition probability make a data assumption?<p>To my ears (eyes?), this discussion resembles the transition from linear, euclidean geometry into the fractal realm.
评论 #19837698 未加载
评论 #19841376 未加载
basetopabout 6 years ago
If people liked this paper, I suggest reading &quot;The Two Cultures&quot; by CP Snow which is not as technical but more expansive, cultural and philosophical.
nestorDabout 6 years ago
&gt; Interpretability is a way of getting information. But a model does not have to be simple to provide reliable information about the relation between predictor and response variables; neither does it have to be a data model. The goal is not interpretability, but accurate information.
vowellessabout 6 years ago
As others have said, great paper by a great author. Must read.
FabHKabout 6 years ago
(2001)