Improving Self-Organizing Maps with Unsupervised Feature Extraction
This work addresses the challenge of improving SOM efficiency for unsupervised learning in embedded applications, though it is incremental as it builds on existing feature extraction methods.
The paper tackles the problem of Self-Organizing Maps (SOMs) performing poorly on complex datasets by using unsupervised feature extraction, resulting in a +6.09% improvement in classification accuracy and achieving state-of-the-art performance on unsupervised image classification.
The Self-Organizing Map (SOM) is a brain-inspired neural model that is very promising for unsupervised learning, especially in embedded applications. However, it is unable to learn efficient prototypes when dealing with complex datasets. We propose in this work to improve the SOM performance by using extracted features instead of raw data. We conduct a comparative study on the SOM classification accuracy with unsupervised feature extraction using two different approaches: a machine learning approach with Sparse Convolutional Auto-Encoders using gradient-based learning, and a neuroscience approach with Spiking Neural Networks using Spike Timing Dependant Plasticity learning. The SOM is trained on the extracted features, then very few labeled samples are used to label the neurons with their corresponding class. We investigate the impact of the feature maps, the SOM size and the labeled subset size on the classification accuracy using the different feature extraction methods. We improve the SOM classification by +6.09\% and reach state-of-the-art performance on unsupervised image classification.