Statistical Mechanics of Time Domain Ensemble Learning

TL;DR

This paper introduces and analyzes time domain ensemble learning, a novel approach that combines students over time rather than space, showing it to be twice as effective as traditional methods in online learning scenarios.

Contribution

The paper presents a new framework for ensemble learning in the time domain and provides a statistical mechanical analysis of its generalization performance.

Findings

Time domain ensemble learning is twice as effective as space domain ensemble learning.

The analysis is conducted within an online learning framework with linear perceptrons.

The approach demonstrates improved generalization performance in noisy environments.

Abstract

Conventional ensemble learning combines students in the space domain. On the other hand, in this paper we combine students in the time domain and call it time domain ensemble learning. In this paper, we analyze the generalization performance of time domain ensemble learning in the framework of online learning using a statistical mechanical method. We treat a model in which both the teacher and the student are linear perceptrons with noises. Time domain ensemble learning is twice as effective as conventional space domain ensemble learning.