[Updated with Andy's Holdout reuse and John Mount's blog entries at the very end of this post, h/t Charles and Giuseppe and WinVector on Twitter]
Over the years, people like Andrew Gelman have been leading the good fight of pointing how some studies, particularly in social sciences and other areas, are marred with issues stemming in part from "p-value hacking" i.e; the ability for poorly designed experiments and inference methods to pick a non relevant set of explanatory variables (see also the Freedman's paradox)
Within the generic context of compressive sensing, this is related, in part, with the issue of underdeterminedness of systems of equations where an infinite number of solutions exist. Compressive Sensing generally focuses on the ability of finding a solution that is simple enough (i.e. sparse) out of this infinite set of possibilities; Noiseless phase transitions are a way to guide use through the ability to identify those sparse solutions given a certain amount of sampling. In noisy situations, those phase transitions change and sometimes, the solution recoverability is not garanteed. The ability to not find a sparse solution could also rear its ugly head in an adversarial setting and/or some adaptive compressive sensing situations.
In Machine Learning, which really extends traditional statistical analysis, the issue is becoming central when evaluating models through the use of the test set. In effect, once you have used the test set, your next iteration is embedding that information to improve its next iteration. Kaggle and Kaggle-like competitions have been set up to deal with this issue. A public leaderboard tells you how your model fits the data with an unknown but reused datasets. A second set of data is then used only once (or secretly) to assess the real position of the algorithms with respect to others. The recent "scandal" in Machine Learning is linked to this ability to reuse the test set more often than the rest of the community. But really deep down, one wonders how often is often. This is why any clever way to reuse the test set is becoming a very interesting proposition.
To get more insight on this issue and how it may be solved, you want to read both of these blog entries and their attendant comments:
- The reusable holdout: Preserving validity in adaptive data analysis by Moritz Hardt
- Gay gene tabloid hype update ( and earlier Latest gay gene tabloid hype ) by Andrew Gelman.
and more importantly the attendant NIPS submission related to the first blog entry:
Generalization in Adaptive Data Analysis and Holdout Reuse by Cynthia Dwork, Vitaly Feldman, Moritz Hardt, Toniann Pitassi, Omer Reingold, Aaron Roth
Overfitting is the bane of data analysts, even when data are plentiful. Formal approaches to understanding this problem focus on statistical inference and generalization of individual analysis procedures. Yet the practice of data analysis is an inherently interactive and adaptive process: new analyses and hypotheses are proposed after seeing the results of previous ones, parameters are tuned on the basis of obtained results, and datasets are shared and reused. An investigation of this gap has recently been initiated by the authors in (Dwork et al., 2014), where we focused on the problem of estimating expectations of adaptively chosen functions.
In this paper, we give a simple and practical method for reusing a holdout (or testing) set to validate the accuracy of hypotheses produced by a learning algorithm operating on a training set. Reusing a holdout set adaptively multiple times can easily lead to overfitting to the holdout set itself. We give an algorithm that enables the validation of a large number of adaptively chosen hypotheses, while provably avoiding overfitting. We illustrate the advantages of our algorithm over the standard use of the holdout set via a simple synthetic experiment.
We also formalize and address the general problem of data reuse in adaptive data analysis. We show how the differential-privacy based approach given in (Dwork et al., 2014) is applicable much more broadly to adaptive data analysis. We then show that a simple approach based on description length can also be used to give guarantees of statistical validity in adaptive settings. Finally, we demonstrate that these incomparable approaches can be unified via the notion of approximate max-information that we introduce.
also of interest: Science's The reusable holdout: Preserving validity in adaptive data analysis by Cynthia Dwork, Vitaly Feldman, Moritz Hardt, Toniann Pitassi, Omer Reingold, Aaron Roth
and STOC 2015's Preserving Statistical Validity in Adaptive Data Analysis. Joint work with Cynthia Dwork, Vitaly Feldman, Moritz Hardt, Toniann Pitassi, Omer Reingold, Aaron Roth
Update:
and STOC 2015's Preserving Statistical Validity in Adaptive Data Analysis. Joint work with Cynthia Dwork, Vitaly Feldman, Moritz Hardt, Toniann Pitassi, Omer Reingold, Aaron Roth
Update:
- Andy has a blog post on the use of this holdout reuse technique in Holdout reuse and attendant Reddit comments.
- John Mount wrote on the Win-vector blog about Using differential privacy to reuse training data
- Nina Zumel, A Simpler Explanation of Differential Privacy
- Moritz's slides
- Reddit comments on the Google Blog entry.
also from the comment section:
Note that the NIPS 2015 paper relies on results from two papers by other authors that came after the STOC 2015 paper:
Bassily et al: http://arxiv.org/abs/1503.04843
Nissim and Stemmer: http://arxiv.org/abs/1504.05800
Those two papers provide tighter generalization guarantees than the STOC 2015 paper. The NIPS 2015 and Science papers rely on those intermediate results.
— charles (@charles_yesman) October 13, 2015
h/t Giuseppe on Twitter
@F_Vaggi @andyjonescs @IgorCarron there was a longer post on win-vector blog too. Procedure is clear. I haven't really read the paper.
— 3pisdn (@gappy3000) October 15, 2015
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.
1 comment:
Note that the NIPS 2015 paper relies on results from two papers by other authors that came after the STOC 2015 paper:
Bassily et al: http://arxiv.org/abs/1503.04843
Nissim and Stemmer: http://arxiv.org/abs/1504.05800
Those two papers provide tighter generalization guarantees than the STOC 2015 paper. The NIPS 2015 and Science papers rely on those intermediate results.
Post a Comment