We transform the outputs of each hidden neuron in a multi-layer perceptron network to have zero output and zero slope on average, and use separate shortcut connections to model th...
In this paper we present a method for learning classspecific
features for recognition. Recently a greedy layerwise
procedure was proposed to initialize weights of deep
belief ne...
Mohammad Norouzi (Simon Fraser University), Mani R...
Abstract. We introduce a concept of self-organizing Hybrid Neurofuzzy Networks (HNFN), a hybrid modeling architecture combining neurofuzzy (NF) and polynomial neural networks(PNN)....
A data simulator that can facilitate the development of improved sampling and analysis procedures for spatial analysis is proposed. The simulator, implemented in MATLAB, provides ...
– This paper presents a hippocampal inspired robot localization model that provides a means for a simple robotic platform with ultrasonic sensors to localize itself. There have b...