Sunday, December 18, 2011

Thoughts on ICDM I: Negative Results (part A)

I just got back from ICDM (the IEEE conference on Data Mining). Data mining conferences are quite different from theory conferences (and much more similar to ML or DB conferences): there are numerous satellite events (workshops, tutorials and panels in this case), many more people (551 for ICDM, and that's on the smaller side), and a wide variety of papers that range from SODA-ish results to user studies and industrial case studies.

While your typical data mining paper is still a string of techniques cobbled together without rhyme or reason (anyone for spectral manifold-based correlation clustering with outliers using MapReduce?), there are some general themes that might be of interest to an outside viewer. What I'd like to highlight here is a trend (that I hope grows) in negative results.

It's not particularly hard to invent a new method for doing data mining. It's much harder to show why certain methods will fail, or why certain models don't make sense. But in my view, the latter is exactly what the field needs in order to give it a strong inferential foundation to build on (I'll note here that I'm talking specifically about data mining, NOT machine learning - the difference between the two is left for another post).

In the interest of brevity, I'll break this trend down in three posts. The first result I want to highlight isn't even quite a result, and isn't even from ICDM ! It's actually from KDD 2011 (back in August this year). The paper is Leakage in Data Mining: Formulation, Detection, and Avoidance, by Shachar Kaufman, Saharon Rosset, and Claudia Perlich, and got the Best Paper Award at KDD this year.

The problem they examine is "leakage", or the phenomenon that even in a 'train model and test model" framework, it is possible for valuable information to "leak" from the test data or even from other sources to the training system, making a learned model look surprisingly effective and even give it predictive powers beyond what it really can do. Obviously, the problem is that when such models are then applied to new data, their performance is worse than expected, compared to a model that wasn't "cheating" in this way.

They cite a number of examples, including many that come from the data challenges that have become all the rage. The examples highlight different kinds of leakage, including "seeing the future", cross contamination of data from other data sets, and even leakage by omission, where a well-intentioned process of anonymization actually leaks data about the trend being analyzed.

While there are no results of any kind (experimental or theoretical), the authors lay out a good taxonomy of common ways in which leakage can happen, and describe ways of avoiding leakage (when you have control over the data) and detecting it (when you don't). What makes their paper really strong is that they illustrate this with specific examples from recent data challenges, explaining how the leakage occurs, and how the winners took advantage of this leakage explicitly or implicitly.


There are no quick and dirty fixes for these problems: ultimately, leakage can even happen through bad modelling, and sometimes modellers fail to remove leaky data because they're trying to encourage competitors to build good predictive models. Ironically, it is this encouragement that can lead to less predictive models on truly novel data. But the paper makes a strong argument that the way we collect data and use it to analyze our algorithms is fundamentally flawed, and this is especially true for the more sophisticated (and mysterious) algorithms that might be learning models through complex exploitation of the trained data.

It remains to be seen whether the recommendations of this paper will be picked up, or if there will be more followup work along these lines. I hope it does.

3 comments:

  1. We need to fix the reviewing process.

    Reviewers often seem to look only at keywords, numbers and charts; not at the involved methods: "Wow, it has MapReduce and outperforms everything else, with 100% accuracy!"
    and don't bother to critically think about whether the method really works.

    These days, way too many papers get accepted just because they present themselves prettily (often even with fake results).

    ReplyDelete
  2. There are journals dedicated exclusively to publish negative results, like The All Results Journals family, where everybody can publish their negative results for free (from their website "the first Total Open Access journal..."); I like this concept of free to publish and read journal ;)

    ReplyDelete
  3. Also related, there are some DM/ML "anti-patterns" listed in Clever Methods of Overfitting.

    ReplyDelete

Disqus for The Geomblog