An Unsupervised Clustering Algorithm Based On Sensory Neuronal Selectivity

Abstract: In this talk I present an unsupervised clustering algorithm that makes use a single neuron to perform classification tasks on a given dataset. The algorithm makes use of the so-called BCM theory of synaptic plasticity, first proposed by Elie Bienenstock, Leon Cooper, and Paul Munro to measure the selectivity of neurons in primary visual cortex and its dependency on neuronal inputs. I will discuss the algorithm, its computational complexity, and how it performs in comparison to bench mark algorithms that perform similar tasks; while paying attention to the qualitative behavior of an underlying differential equation.