An official website of the United States government
Data Valuation in machine learning is concerned with quantifying the relative contribution of a training example to a model’s performance. Quantifying the importance of training examples is useful for identifying high and low quality data to curate training datasets and for address data quality issues. Shapley values have gained traction in machine learning for curating training data and identifying data quality issues. While computing the Shapley values of training examples is computationally prohibitive, approximation methods have been used successfully for classification models in computer vision tasks. We investigate data valuation for Automatic Speech Recognition models which perform a structured prediction task and propose a method for estimating Shapley values for these models. We show that a proxy model can be learned for the acoustic model component of an end-to-end ASR and used to estimate Shapley values for acoustic frames. We present a method for using the proxy acoustic model to estimate Shapley values for variable length utterances and demonstrate that the Shapley values provide a signal of example quality.
more »
« less
FastSHAP: Real-Time Shapley Value Estimation
Although Shapley values are theoretically appealing for explaining black-box models, they are costly to calculate and thus impractical in settings that involve large, high-dimensional models. To remedy this issue, we introduce FastSHAP, a new method for estimating Shapley values in a single forward pass using a learned explainer model. To enable efficient training without requiring ground truth Shapley values, we develop an approach to train FastSHAP via stochastic gradient descent using a weighted least squares objective function. In our experiments with tabular and image datasets, we compare FastSHAP to existing estimation approaches and f ind that it generates accurate explanations with an orders-of-magnitude speedup.
more »
« less
- Award ID(s):
- 1922658
- PAR ID:
- 10350825
- Date Published:
- Journal Name:
- ICLR 2022
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Popular feature importance techniques compute additive approximations to nonlinear models by first defining a cooperative game describing the value of different subsets of the model’s features, then calculating the resulting game’s Shapley values to attribute credit additively between the features. However, the specific modeling settings in which the Shapley values are a poor approximation for the true game have not been well-described. In this paper we utilize an interpretation of Shapley values as the result of an orthogonal projection between vector spaces to calculate a residual representing the kernel component of that projection. We provide an algorithm for computing these residuals, characterize different modeling settings based on the value of the residuals, and demonstrate that they capture information about model predictions that Shapley values cannot.more » « less
-
Popular feature importance techniques compute additive approximations to nonlinear models by first defining a cooperative game describing the value of different subsets of the model’s features, then calculating the resulting game’s Shapley values to attribute credit additively between the features. However, the specific modeling settings in which the Shapley values are a poor approximation for the true game have not been well-described. In this paper we utilize an interpretation of Shapley values as the result of an orthogonal projection between vector spaces to calculate a residual representing the kernel component of that projection. We provide an algorithm for computing these residuals, characterize different modeling settings based on the value of the residuals, and demonstrate that they capture information about model predictions that Shapley values cannot.more » « less
-
Shapley value is a classic notion from game theory, historically used to quantify the contributions of individuals within groups, and more recently applied to assign values to data points when training machine learning models. Despite its foundational role, a key limitation of the data Shapley framework is that it only provides valuations for points within a fixed data set. It does not account for statistical aspects of the data and does not give a way to reason about points outside the data set. To address these limitations, we propose a novel framework -- distributional Shapley -- where the value of a point is defined in the context of an underlying data distribution. We prove that distributional Shapley has several desirable statistical properties; for example, the values are stable under perturbations to the data points themselves and to the underlying data distribution. We leverage these properties to develop a new algorithm for estimating values from data, which comes with formal guarantees and runs two orders of magnitude faster than state-of-the-art algorithms for computing the (non-distributional) data Shapley values. We apply distributional Shapley to diverse data sets and demonstrate its utility in a data market setting.more » « less
-
In this paper we investigate the problem of quantifying the contribution of each variable to the satisfying assignments of a Boolean function based on the Shapley value. Our main result is a polynomial-time equivalence between computing Shapley values and model counting for any class of Boolean functions that are closed under substitutions of variables with disjunctions of fresh variables. This result settles an open problem raised in prior work, which sought to connect the Shapley value computation to probabilistic query evaluation. We show two applications of our result. First, the Shapley values can be computed in polynomial time over deterministic and decomposable circuits, since they are closed under OR-substitutions. Second, there is a polynomial-time equivalence between computing the Shapley value for the tuples contributing to the answer of a Boolean conjunctive query and counting the models in the lineage of the query. This equivalence allows us to immediately recover the dichotomy for Shapley value computation in case of self-join-free Boolean conjunctive queries; in particular, the hardness for non-hierarchical queries can now be shown using a simple reduction from the \#P-hard problem of model counting for lineage in positive bipartite disjunctive normal form.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
