Iterative Domain Optimization

TL;DR

This paper introduces an iterative gradient-based optimization method to identify large, small, or specific value domains of a function, with applications in machine learning model output analysis, demonstrated on Titanic dataset models.

Contribution

The paper presents a novel iterative optimization approach that approximates a non-optimizable objective function, enabling domain search for model outputs under constraints.

Findings

Effective in identifying output domains in five case studies

Demonstrates efficiency on models trained on Titanic dataset

Provides a new tool for model output analysis

Abstract

In this paper we study a new approach in optimization that aims to search a large domain D where a given function takes large, small or specific values via an iterative optimization algorithm based on the gradient. We show that the objective function used is not directly optimizable, however, we use a trick to approximate this objective by another one at each iteration to optimize it. Then we explore a use case of this algorithm in machine learning to find domains where the models output large and small values with respect of some constraints. Experiments demonstrate the efficiency of this algorithm on five cases with models trained on the titanic dataset.