Theoretical Behavior of XAI Methods in the Presence of Suppressor Variables

• URL: http://arxiv.org/abs/2306.01464v1
• Date: Fri, 2 Jun 2023 11:41:19 GMT
• Title: Theoretical Behavior of XAI Methods in the Presence of Suppressor Variables
• Authors: Rick Wilming, Leo Kieslich, Benedict Clark, Stefan Haufe
• Abstract summary: In recent years, the community of 'explainable artificial intelligence' (XAI) has created a vast body of methods to bridge a perceived gap between model 'complexity' and 'interpretability' We show that the majority of the studied approaches will attribute non-zero importance to a non-class-related suppressor feature in the presence of correlated noise.
• Score: 0.8602553195689513
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: In recent years, the community of 'explainable artificial intelligence' (XAI) has created a vast body of methods to bridge a perceived gap between model 'complexity' and 'interpretability'. However, a concrete problem to be solved by XAI methods has not yet been formally stated. As a result, XAI methods are lacking theoretical and empirical evidence for the 'correctness' of their explanations, limiting their potential use for quality-control and transparency purposes. At the same time, Haufe et al. (2014) showed, using simple toy examples, that even standard interpretations of linear models can be highly misleading. Specifically, high importance may be attributed to so-called suppressor variables lacking any statistical relation to the prediction target. This behavior has been confirmed empirically for a large array of XAI methods in Wilming et al. (2022). Here, we go one step further by deriving analytical expressions for the behavior of a variety of popular XAI methods on a simple two-dimensional binary classification problem involving Gaussian class-conditional distributions. We show that the majority of the studied approaches will attribute non-zero importance to a non-class-related suppressor feature in the presence of correlated noise. This poses important limitations on the interpretations and conclusions that the outputs of these XAI methods can afford.

Related papers

• Assouad, Fano, and Le Cam with Interaction: A Unifying Lower Bound Framework and Characterization for Bandit Learnability [71.82666334363174]
  We develop a unified framework for lower bound methods in statistical estimation and interactive decision making. We introduce a novel measure, decision dimension, which facilitates the complexity of new lower bounds for interactive decision making.
  arXiv  Detail & Related papers  (2024-10-07T15:14:58Z)
• Unified Explanations in Machine Learning Models: A Perturbation Approach [0.0]
  Inconsistencies between XAI and modeling techniques can have the undesirable effect of casting doubt upon the efficacy of these explainability approaches. We propose a systematic, perturbation-based analysis against a popular, model-agnostic method in XAI, SHapley Additive exPlanations (Shap) We devise algorithms to generate relative feature importance in settings of dynamic inference amongst a suite of popular machine learning and deep learning methods, and metrics that allow us to quantify how well explanations generated under the static case hold.
  arXiv  Detail & Related papers  (2024-05-30T16:04:35Z)
• XpertAI: uncovering model strategies for sub-manifolds [1.2874569408514918]
  In regression, explanations need to be precisely formulated to address specific user queries. We introduce XpertAI, a framework that disentangles the prediction strategy into multiple range-specific sub-strategies.
  arXiv  Detail & Related papers  (2024-03-12T10:21:31Z)
• XAI-TRIS: Non-linear image benchmarks to quantify false positive post-hoc attribution of feature importance [1.3958169829527285]
  A lack of formal underpinning leaves it unclear as to what conclusions can safely be drawn from the results of a given XAI method. This means that challenging non-linear problems, typically solved by deep neural networks, presently lack appropriate remedies. We show that popular XAI methods are often unable to significantly outperform random performance baselines and edge detection methods.
  arXiv  Detail & Related papers  (2023-06-22T11:31:11Z)
• Explaining Explainability: Towards Deeper Actionable Insights into Deep Learning through Second-order Explainability [70.60433013657693]
  Second-order explainable AI (SOXAI) was recently proposed to extend explainable AI (XAI) from the instance level to the dataset level. We demonstrate for the first time, via example classification and segmentation cases, that eliminating irrelevant concepts from the training set based on actionable insights from SOXAI can enhance a model's performance.
  arXiv  Detail & Related papers  (2023-06-14T23:24:01Z)
• Characterizing the contribution of dependent features in XAI methods [6.990173577370281]
  We propose a proxy that modifies the outcome of any XAI feature ranking method allowing to account for the dependency among the predictors. The proposed approach has the advantage of being model-agnostic as well as simple to calculate the impact of each predictor in the model in presence of collinearity.
  arXiv  Detail & Related papers  (2023-04-04T11:25:57Z)
• Optimizing Explanations by Network Canonization and Hyperparameter Search [74.76732413972005]
  Rule-based and modified backpropagation XAI approaches often face challenges when being applied to modern model architectures. Model canonization is the process of re-structuring the model to disregard problematic components without changing the underlying function. In this work, we propose canonizations for currently relevant model blocks applicable to popular deep neural network architectures.
  arXiv  Detail & Related papers  (2022-11-30T17:17:55Z)
• Principled Knowledge Extrapolation with GANs [92.62635018136476]
  We study counterfactual synthesis from a new perspective of knowledge extrapolation. We show that an adversarial game with a closed-form discriminator can be used to address the knowledge extrapolation problem. Our method enjoys both elegant theoretical guarantees and superior performance in many scenarios.
  arXiv  Detail & Related papers  (2022-05-21T08:39:42Z)
• Estimation of Bivariate Structural Causal Models by Variational Gaussian Process Regression Under Likelihoods Parametrised by Normalising Flows [74.85071867225533]
  Causal mechanisms can be described by structural causal models. One major drawback of state-of-the-art artificial intelligence is its lack of explainability.
  arXiv  Detail & Related papers  (2021-09-06T14:52:58Z)
• Rational Shapley Values [0.0]
  Most popular tools for post-hoc explainable artificial intelligence (XAI) are either insensitive to context or difficult to summarize. I introduce emphrational Shapley values, a novel XAI method that synthesizes and extends these seemingly incompatible approaches. I leverage tools from decision theory and causal modeling to formalize and implement a pragmatic approach that resolves a number of known challenges in XAI.
  arXiv  Detail & Related papers  (2021-06-18T15:45:21Z)
• Good Classifiers are Abundant in the Interpolating Regime [64.72044662855612]
  We develop a methodology to compute precisely the full distribution of test errors among interpolating classifiers. We find that test errors tend to concentrate around a small typical value $varepsilon*$, which deviates substantially from the test error of worst-case interpolating model. Our results show that the usual style of analysis in statistical learning theory may not be fine-grained enough to capture the good generalization performance observed in practice.
  arXiv  Detail & Related papers  (2020-06-22T21:12:31Z)

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.