Online Sliding Window Based Self-Organising Fuzzy Neural Network for Cognitive Reasoning
Gang Leng, Anjan Kumar Ray, T.M. McGinnity, Sonya Coleman, Liam Maguire · 2013
Abstract —We propose an online sliding window based self-organising fuzzy neural network (SOFNN) as the core component of a cognitive reasoning system for a smart home environment. The network has the ability to configure its neuronal structure through adding and pruning of neurons while exploring the relationships between the inputs and the desired reasoning outputs, thus enabling continuous learning and reasoning to provide meaningful cognitive understanding of the environment. Initially, the network is trained with environmentally realistic synthesised data thus demonstrating its adaptation capabilities. The network is then validated using unseen data. In the simulation, we have studied the network structures and responses for three different scenarios with and without online sliding window based approaches and the results obtained show the effectiveness of the proposed method. Keywords- self-organise; fuzzy logic; neural network; reasoning module I.