Learning Sign-Constrained Support Vector Machines

TL;DR

This paper introduces two optimization algorithms for training linear support vector machines with sign constraints on weights, leveraging domain knowledge to potentially improve generalization, and provides theoretical convergence analysis and empirical validation.

Contribution

It develops and analyzes two efficient algorithms for sign-constrained SVMs, incorporating domain knowledge into the learning process with proven convergence guarantees.

Findings

Algorithms converge sublinearly with $O(nd)$ per iteration cost

Explicit minimal iteration number for $psilon$-accurate solutions derived

Empirical results show sign constraints enhance performance with similarity-based features

Abstract

Domain knowledge is useful to improve the generalization performance of learning machines. Sign constraints are a handy representation to combine domain knowledge with learning machine. In this paper, we consider constraining the signs of the weight coefficients in learning the linear support vector machine, and develop two optimization algorithms for minimizing the empirical risk under the sign constraints. One of the two algorithms is based on the projected gradient method, in which each iteration of the projected gradient method takes $O(nd)$ computational cost and the sublinear convergence of the objective error is guaranteed. The second algorithm is based on the Frank-Wolfe method that also converges sublinearly and possesses a clear termination criterion. We show that each iteration of the Frank-Wolfe also requires $O(nd)$ cost. Furthermore, we derive the explicit expression for the minimal iteration number to ensure an $\epsilon$-accurate solution by analyzing the curvature of the objective function. Finally, we empirically demonstrate that the sign constraints are a promising technique when similarities to the training examples compose the feature vector.