Kohonen's Self-Organizing Feature Maps and Linear Vector Quantization: A Comparison

Machine learning has evolved over the past years to become one of the major research fields in Computer Science. In simple words, Machine Learning can be described as the process of training a machine to learn from its outputs and improvise itself in order to optimize its outputs. One of the major branch of machine learning is Unsupervised Learning where in the machine is not given any kind of feedback but is expected to learn on its own (“without Supervision”). This paper aims at describing in detail and thus comparing two such neural networks: Kohonen’s Self Organizing Feature Maps (KSOFM) and Linear Vector Quantization (LVQ).