Multi-Model Least Squares-Based Recomputation Framework for Large Data Analysis

TL;DR

This paper introduces a recomputation framework for multilayer least squares neural networks that enhances generalization by retraining hidden layers using Moore-Penrose inverse, demonstrated on large datasets.

Contribution

It proposes a novel multilayer least squares network with a retraining strategy using Moore-Penrose inverse, improving performance on large-scale data.

Findings

Better generalization performance than most representation learning algorithms.

Effective on datasets ranging from 3K to 1.8M samples.

Sparse RML-MP boosts the performance of the basic RML-MP.

Abstract

Most multilayer least squares (LS)-based neural networks are structured with two separate stages: unsupervised feature encoding and supervised pattern classification. Once the unsupervised learning is finished, the latent encoding would be fixed without supervised fine-tuning. However, in complex tasks such as handling the ImageNet dataset, there are often many more clues that can be directly encoded, while the unsupervised learning, by definition cannot know exactly what is useful for a certain task. This serves as the motivation to retrain the latent space representations to learn some clues that unsupervised learning has not yet learned. In particular, the error matrix from the output layer is pulled back to each hidden layer, and the parameters of the hidden layer are recalculated with Moore-Penrose (MP) inverse for more generalized representations. In this paper, a recomputation-based multilayer network using MP inverse (RML-MP) is developed. A sparse RML-MP (SRML-MP) model to boost the performance of RML-MP is then proposed. The experimental results with varying training samples (from 3 K to 1.8 M) show that the proposed models provide better generalization performance than most representation learning algorithms.