On Correcting SHAP Scores

• URL: http://arxiv.org/abs/2405.00076v1
• Date: Tue, 30 Apr 2024 10:39:20 GMT
• Title: On Correcting SHAP Scores
• Authors: Olivier Letoffe, Xuanxiang Huang, Joao Marques-Silva,
• Abstract summary: The paper makes the case that the failings of SHAP scores result from the characteristic functions used in earlier works. The paper proposes modifications to the tool SHAP to use instead one of our novel characteristic functions.
• Score: 3.3766484312332303
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Recent work uncovered examples of classifiers for which SHAP scores yield misleading feature attributions. While such examples might be perceived as suggesting the inadequacy of Shapley values for explainability, this paper shows that the source of the identified shortcomings of SHAP scores resides elsewhere. Concretely, the paper makes the case that the failings of SHAP scores result from the characteristic functions used in earlier works. Furthermore, the paper identifies a number of properties that characteristic functions ought to respect, and proposes several novel characteristic functions, each exhibiting one or more of the desired properties. More importantly, some of the characteristic functions proposed in this paper are guaranteed not to exhibit any of the shortcomings uncovered by earlier work. The paper also investigates the impact of the new characteristic functions on the complexity of computing SHAP scores. Finally, the paper proposes modifications to the tool SHAP to use instead one of our novel characteristic functions, thereby eliminating some of the limitations reported for SHAP scores.

Related papers

• From SHAP Scores to Feature Importance Scores [4.8158930873043335]
  This paper shows that there is an essential relationship between feature attribution and a priori voting power. It remains unclear how some of the most widely used power indices might be exploited as feature importance scores (FISs) in XAI.
  arXiv  Detail & Related papers  (2024-05-20T03:52:41Z)
• Spurious Feature Eraser: Stabilizing Test-Time Adaptation for Vision-Language Foundation Model [86.9619638550683]
  Vision-language foundation models have exhibited remarkable success across a multitude of downstream tasks due to their scalability on extensive image-text paired data. However, these models display significant limitations when applied to downstream tasks, such as fine-grained image classification, as a result of decision shortcuts''
  arXiv  Detail & Related papers  (2024-03-01T09:01:53Z)
• Keypoint Description by Symmetry Assessment -- Applications in Biometrics [49.547569925407814]
  We present a model-based feature extractor to describe neighborhoods around keypoints by finite expansion. The iso-curves of such functions are highly symmetric w.r.t. the origin (a keypoint) and the estimated parameters have well defined geometric interpretations.
  arXiv  Detail & Related papers  (2023-11-03T00:49:25Z)
• A Refutation of Shapley Values for Explainability [4.483306836710804]
  Recent work demonstrated the existence of Boolean functions for which Shapley values provide misleading information. This paper proves that, for any number of features, there exist Boolean functions that exhibit one or more inadequacy-revealing issues.
  arXiv  Detail & Related papers  (2023-09-06T14:34:18Z)
• Regularization Trade-offs with Fake Features [0.0]
  This paper considers a framework where the possibly overparametrized model includes fake features. We present a non-asymptotic high-probability bound on the generalization error of the ridge regression problem.
  arXiv  Detail & Related papers  (2022-12-01T11:11:14Z)
• A Fair Loss Function for Network Pruning [93.0013343535411]
  We introduce the performance weighted loss function, a simple modified cross-entropy loss function that can be used to limit the introduction of biases during pruning. Experiments using biased classifiers for facial classification and skin-lesion classification tasks demonstrate that the proposed method is a simple and effective tool.
  arXiv  Detail & Related papers  (2022-11-18T15:17:28Z)
• Inference on Strongly Identified Functionals of Weakly Identified Functions [71.42652863687117]
  We study a novel condition for the functional to be strongly identified even when the nuisance function is not. We propose penalized minimax estimators for both the primary and debiasing nuisance functions.
  arXiv  Detail & Related papers  (2022-08-17T13:38:31Z)
• Causal Inference Under Unmeasured Confounding With Negative Controls: A Minimax Learning Approach [84.29777236590674]
  We study the estimation of causal parameters when not all confounders are observed and instead negative controls are available. Recent work has shown how these can enable identification and efficient estimation via two so-called bridge functions.
  arXiv  Detail & Related papers  (2021-03-25T17:59:19Z)
• Towards Unifying Feature Attribution and Counterfactual Explanations: Different Means to the Same End [17.226134854746267]
  We present a method to generate feature attribution explanations from a set of counterfactual examples. We show how counterfactual examples can be used to evaluate the goodness of an attribution-based explanation in terms of its necessity and sufficiency.
  arXiv  Detail & Related papers  (2020-11-10T05:41:43Z)
• An Equivalence between Loss Functions and Non-Uniform Sampling in Experience Replay [72.23433407017558]
  We show that any loss function evaluated with non-uniformly sampled data can be transformed into another uniformly sampled loss function. Surprisingly, we find in some environments PER can be replaced entirely by this new loss function without impact to empirical performance.
  arXiv  Detail & Related papers  (2020-07-12T17:45:24Z)
• Interpretable feature subset selection: A Shapley value based approach [1.511944009967492]
  We introduce the notion of classification game, a cooperative game with features as players and hinge loss based characteristic function. Our major contribution is ($star$) to show that for any dataset the threshold 0 on SVEA value identifies feature subset whose joint interactions for label prediction is significant.
  arXiv  Detail & Related papers  (2020-01-12T16:27:08Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.