An Unsupervised Learning Classifier with Competitive Error Performance

TL;DR

This paper introduces an unsupervised classifier that rivals supervised methods like SVM and kNN in error performance by iteratively updating hyperplanes, demonstrating competitive results on ImageNet with a simple feature extractor.

Contribution

The paper presents a novel unsupervised learning classifier based on incremental hyperplane adjustments, achieving error rates close to supervised classifiers on large-scale image data.

Findings

Achieves 6.2% Top 3 error on ImageNet subset

Outperforms k-Means in unsupervised classification

Comparable to kNN and SVM in error probability

Abstract

An unsupervised learning classification model is described. It achieves classification error probability competitive with that of popular supervised learning classifiers such as SVM or kNN. The model is based on the incremental execution of small step shift and rotation operations upon selected discriminative hyperplanes at the arrival of input samples. When applied, in conjunction with a selected feature extractor, to a subset of the ImageNet dataset benchmark, it yields 6.2 % Top 3 probability of error; this exceeds by merely about 2 % the result achieved by (supervised) k-Nearest Neighbor, both using same feature extractor. This result may also be contrasted with popular unsupervised learning schemes such as k-Means which is shown to be practically useless on same dataset.