The purpose of this study is to develop a flexible matching method for recognizing handwritten numerals based on the statistics of shapes and structures learned from learning sam...
In this paper we propose a Neural Net-PMRS hybrid for forecasting time-series data. The neural network model uses the traditional MLP architecture and backpropagation method of tr...
In this paper we describe a heuristic approach to the problem of identifying a pattern embedded within a figure from a predefined set of patterns via the utilization of a genetic a...
: Linear subspace analysis (LSA) has become rather ubiquitous in a wide range of problems arising in pattern recognition and computer vision. The essence of these approaches is tha...