Deep Categorization with Semi-Supervised Self-Organizing Maps

TL;DR

This paper introduces Batch SS-SOM, a semi-supervised self-organizing map model that effectively leverages both labeled and unlabeled data for classification and clustering, demonstrating competitive results in various scenarios.

Contribution

It presents the Batch SS-SOM, an extension of SOM that incorporates batch training and deep learning advances for semi-supervised learning.

Findings

Performs well with few labeled samples

Achieves low clustering error

Competitive in transfer learning tasks

Abstract

Nowadays, with the advance of technology, there is an increasing amount of unstructured data being generated every day. However, it is a painful job to label and organize it. Labeling is an expensive, time-consuming, and difficult task. It is usually done manually, which collaborates with the incorporation of noise and errors to the data. Hence, it is of great importance to developing intelligent models that can benefit from both labeled and unlabeled data. Currently, works on unsupervised and semi-supervised learning are still being overshadowed by the successes of purely supervised learning. However, it is expected that they become far more important in the longer term. This article presents a semi-supervised model, called Batch Semi-Supervised Self-Organizing Map (Batch SS-SOM), which is an extension of a SOM incorporating some advances that came with the rise of Deep Learning, such as batch training. The results show that Batch SS-SOM is a good option for semi-supervised classification and clustering. It performs well in terms of accuracy and clustering error, even with a small number of labeled samples, as well as when presented to unsupervised data, and shows competitive results in transfer learning scenarios in traditional image classification benchmark datasets.