"How much is my data worth?" is an increasingly common question posed by organizations and individuals alike. An answer to this question could allow, for instance, fairly distributing profits among multiple data contributors and determining prospective compensation when data breaches happen. In this paper, we study the problem of data valuation by utilizing the Shapley value, a popular notion of value which originated in coopoerative game theory. The Shapley value defines a unique payoff scheme that satisfies many desiderata for the notion of data value. However, the Shapley value often requires exponential time to compute. To meet this challenge, we propose a repertoire of efficient algorithms for approximating the Shapley value. We also demonstrate the value of each training instance for various benchmark datasets.

Towards Efficient Data Valuation Based on the Shapley Value.

Given a data set D containing millions of data points and a data consumer who is willing to pay for $X to train a machine learning (ML) model over D, how should we distribute this $X to each data point to reflect its "value"? In this paper, we define the "relative value of data" via the Shapley value, as it uniquely possesses properties with appealing real-world interpretations, such as fairness, rationality and decentralizability. For general, bounded utility functions, the Shapley value is known to be challenging to compute: to get Shapley values for all N data points, it requires O(2N) model evaluations for exact computation and O(N log N) for (ϵ, δ)-approximation.In this paper, we focus on one popular family of ML models relying on K-nearest neighbors (KNN). The most surprising result is that for unweighted KNN classifiers and regressors, the Shapley value of all N data points can be computed, exactly, in O(N log N) time - an exponential improvement on computational complexity! Moreover, for (ϵ, δ)-approximation, we are able to develop an algorithm based on Locality Sensitive Hashing (LSH) with only sublinear complexity O(Nh(ϵ, K) log N) when ϵ is not too small and K is not too large. We empirically evaluate our algorithms on up to 10 million data points and even our exact algorithm is up to three orders of magnitude faster than the baseline approximation algorithm. The LSH-based approximation algorithm can accelerate the value calculation process even further.We then extend our algorithm to other scenarios such as (1) weighed KNN classifiers, (2) different data points are clustered by different data curators, and (3) there are data analysts providing computation who also requires proper valuation. Some of these extensions, although also being improved exponentially, are less practical for exact computation (e.g., O(NK) complexity for weigthed KNN). We thus propose an Monte Carlo approximation algorithm, which is O(N(log N)2/(log K)2) times more efficient than the baseline approximation algorithm.

/pdf/efficient-task-specific-data-valuation-for-nearest-neighbor-26cs0j2drv.pdf

Efficient task-specific data valuation for nearest neighbor algorithms

{\em ``How much is my data worth?''} is an increasingly common question posed by organizations and individuals alike. An answer to this question could allow, for instance, fairly distributing profits among multiple data contributors and determining prospective compensation when data breaches happen. In this paper, we study the problem of \emph{data valuation} by utilizing the Shapley value, a popular notion of value which originated in coopoerative game theory. The Shapley value defines a unique payoff scheme that satisfies many desiderata for the notion of data value. However, the Shapley value often requires \emph{exponential} time to compute.
To meet this challenge, we propose a repertoire of efficient algorithms for approximating the Shapley value. We also demonstrate the value of each training instance for various benchmark datasets.

/pdf/towards-efficient-data-valuation-based-on-the-shapley-value-4fzj0r20a5.pdf

Towards Efficient Data Valuation Based on the Shapley Value

Sparse learning techniques have been routinely used for feature selection as the resulting model usually has a small number of non-zero entries. Safe screening, which eliminates the features that are guaranteed to have zero coefficients for a certain value of the regularization parameter, is a technique for improving the computational efficiency. Safe screening is gaining increasing attention since 1) solving sparse learning formulations usually has a high computational cost especially when the number of features is large and 2) one needs to try several regularization parameters to select a suitable model. In this paper, we propose an approach called "Sasvi" (Safe screening with variational inequalities). Sasvi makes use of the variational inequality that provides the sufficient and necessary optimality condition for the dual problem. Several existing approaches for Lasso screening can be casted as relaxed versions of the proposed Sasvi, thus Sasvi provides a stronger safe screening rule. We further study the monotone properties of Sasvi for Lasso, based on which a sure removal regularization parameter can be identified for each feature. Experimental results on both synthetic and real data sets are reported to demonstrate the effectiveness of the proposed Sasvi for Lasso screening.

/pdf/safe-screening-with-variational-inequalities-and-its-2xuhw6gi1k.pdf

Safe Screening with Variational Inequalities and Its Application to Lasso

Safe Screening With Variational Inequalities and Its Application to LASSO

In this paper, we claim that some of the nonsupport vectors (non-SVs) that have no influence on the SVM classifier can be screened out prior to the training phase in pathwise SVM computation scenario, in which one is asked to train a sequence (or path) of SVM classifiers for different regularization parameters. Based on a recently proposed framework so-called safe screening rule, we derive a rule for screening out non-SVs in advance, and discuss how we can exploit the advantage of the rule in pathwise SVM computation scenario. Experiments indicate that our approach often substantially reduce the total pathwise computation cost.

/pdf/safe-screening-of-non-support-vectors-in-pathwise-svm-37p0u82flw.pdf

Safe Screening of Non-Support Vectors in Pathwise SVM Computation

We introduce a novel sensitivity analysis framework for large scale classification problems that can be used when a small number of instances are incrementally added or removed. For quickly updating the classifier in such a situation, incremental learning algorithms have been intensively studied in the literature. Although they are much more efficient than solving the optimization problem from scratch, their computational complexity yet depends on the entire training set size. It means that, if the original training set is large, completely solving an incremental learning problem might be still rather expensive. To circumvent this computational issue, we propose a novel framework that allows us to make an inference about the updated classifier without actually re-optimizing it. Specifically, the proposed framework can quickly provide a lower and an upper bounds of a quantity on the unknown updated classifier. The main advantage of the proposed framework is that the computational cost of computing these bounds depends only on the number of updated instances. This property is quite advantageous in a typical sensitivity analysis task where only a small number of instances are updated. In this paper we demonstrate that the proposed framework is applicable to various practical sensitivity analysis tasks, and the bounds provided by the framework are often sufficiently tight for making desired inferences.

Quick sensitivity analysis for incremental data modification and its application to leave-one-out CV in linear classification problems

Quick Sensitivity Analysis for Incremental Data Modification and Its Application to Leave-one-out CV in Linear Classification Problems

Sparse classifiers such as the support vector machines (SVM) are efficient in test-phases because the classifier is characterized only by a subset of the samples called support vectors (SVs), and the rest of the samples (non SVs) have no influence on the classification result. However, the advantage of the sparsity has not been fully exploited in training phases because it is generally difficult to know which sample turns out to be SV beforehand. In this paper, we introduce a new approach called safe sample screening that enables us to identify a subset of the non-SVs and screen them out prior to the training phase. Our approach is different from existing heuristic approaches in the sense that the screened samples are guaranteed to be non-SVs at the optimal solution. We investigate the advantage of the safe sample screening approach through intensive numerical experiments, and demonstrate that it can substantially decrease the computational cost of the state-of-the-art SVM solvers such as LIBSVM. In the current big data era, we believe that safe sample screening would be of great practical importance since the data size can be reduced without sacrificing the optimality of the final solution.

Safe Sample Screening for Support Vector Machines

Careful tuning of a regularization parameter is indispensable in many machine learning tasks because it has a significant impact on generalization performances. Nevertheless, current practice of regularization parameter tuning is more of an art than a science, e.g., it is hard to tell how many grid-points would be needed in cross-validation (CV) for obtaining a solution with sufficiently small CV error. In this paper we propose a novel framework for computing a lower bound of the CV errors as a function of the regularization parameter, which we call regularization path of CV error lower bounds. The proposed framework can be used for providing a theoretical approximation guarantee on a set of solutions in the sense that how far the CV error of the current best solution could be away from best possible CV error in the entire range of the regularization parameters. Our numerical experiments demonstrate that a theoretically guaranteed choice of a regularization parameter in the above sense is possible with reasonable computational costs.

/pdf/regularization-path-of-cross-validation-error-lower-bounds-4eohtmsgx2.pdf

Yoshiki Suzuki

Papers

Safe Screening of Non-Support Vectors in Pathwise SVM Computation

Quick sensitivity analysis for incremental data modification and its application to leave-one-out CV in linear classification problems

Quick Sensitivity Analysis for Incremental Data Modification and Its Application to Leave-one-out CV in Linear Classification Problems

Safe Sample Screening for Support Vector Machines

Regularization path of cross-validation error lower bounds