Post Training in Deep Learning with Last Kernel

TL;DR

This paper introduces a post-training method focusing on optimizing only the last layer of deep neural networks, leveraging kernel theory to enhance data representations and improve performance across various architectures and datasets.

Contribution

It proposes a novel post-training step that optimizes only the last layer, analyzed through kernel theory, to improve deep learning performance.

Findings

Consistent performance boost across multiple architectures.

Effective data embedding optimization improves task-specific accuracy.

Kernel-based analysis provides theoretical understanding of the method.

Abstract

One of the main challenges of deep learning methods is the choice of an appropriate training strategy. In particular, additional steps, such as unsupervised pre-training, have been shown to greatly improve the performances of deep structures. In this article, we propose an extra training step, called post-training, which only optimizes the last layer of the network. We show that this procedure can be analyzed in the context of kernel theory, with the first layers computing an embedding of the data and the last layer a statistical model to solve the task based on this embedding. This step makes sure that the embedding, or representation, of the data is used in the best possible way for the considered task. This idea is then tested on multiple architectures with various data sets, showing that it consistently provides a boost in performance.