On the KL-Divergence-based Robust Satisficing Model

TL;DR

This paper introduces a KL-divergence-based robust satisficing model that improves machine learning robustness, especially for deep neural networks, by providing analytical insights, efficient algorithms, and demonstrating superior performance across multiple tasks.

Contribution

It extends the robust satisficing framework to general loss functions, offering analytical interpretations, convergence analysis, and scalable methods for complex models like deep neural networks.

Findings

Outperforms state-of-the-art benchmarks in three machine learning tasks.

Provides stable and efficient numerical algorithms with convergence guarantees.

Extends the model to hierarchical data structures.

Abstract

Empirical risk minimization, a cornerstone in machine learning, is often hindered by the Optimizer's Curse stemming from discrepancies between the empirical and true data-generating distributions.To address this challenge, the robust satisficing framework has emerged recently to mitigate ambiguity in the true distribution. Distinguished by its interpretable hyperparameter and enhanced performance guarantees, this approach has attracted increasing attention from academia. However, its applicability in tackling general machine learning problems, notably deep neural networks, remains largely unexplored due to the computational challenges in solving this model efficiently across general loss functions. In this study, we delve into the Kullback Leibler divergence based robust satisficing model under a general loss function, presenting analytical interpretations, diverse performance guarantees, efficient and stable numerical methods, convergence analysis, and an extension tailored for hierarchical data structures. Through extensive numerical experiments across three distinct machine learning tasks, we demonstrate the superior performance of our model compared to state-of-the-art benchmarks.