{
  "id": "1907.03419",
  "version": "v1",
  "published": "2019-07-08T06:42:59.000Z",
  "updated": "2019-07-08T06:42:59.000Z",
  "title": "The Price of Interpretability",
  "authors": [
    "Dimitris Bertsimas",
    "Arthur Delarue",
    "Patrick Jaillet",
    "Sebastien Martin"
  ],
  "categories": [
    "cs.LG",
    "stat.ML"
  ],
  "abstract": "When quantitative models are used to support decision-making on complex and important topics, understanding a model's ``reasoning'' can increase trust in its predictions, expose hidden biases, or reduce vulnerability to adversarial attacks. However, the concept of interpretability remains loosely defined and application-specific. In this paper, we introduce a mathematical framework in which machine learning models are constructed in a sequence of interpretable steps. We show that for a variety of models, a natural choice of interpretable steps recovers standard interpretability proxies (e.g., sparsity in linear models). We then generalize these proxies to yield a parametrized family of consistent measures of model interpretability. This formal definition allows us to quantify the ``price'' of interpretability, i.e., the tradeoff with predictive accuracy. We demonstrate practical algorithms to apply our framework on real and synthetic datasets.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2019-07-08T06:42:59.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "expose hidden biases",
      "interpretable steps",
      "standard interpretability proxies",
      "reduce vulnerability",
      "adversarial attacks"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}