International Journal of Computer Science & Engineering Technology (IJCSET)
In this paper, the authors present a new method that provides quick convergence for a feed forward neural networks system using feature selection and weight optimization techniques. A neural networks system functions on the basis of weights presented to the neurons. These weights are fine tuned to the most possible accurate level using the training data set. But this process is time and resource intensive, and increases exponentially with the increase in number of attributes.