Prediction of Monthly Min & Max Stock Prices using Neural Network & Genetic Algorithm Hybrid

This paper proposes a hybrid model of Backpropagation (BP) & Genetic Algorithm (GA) to prediction of monthly min and max values of stock prices using historical daily stock prices. Variance in stock price is nonlinear and one of the best known approaches in field of Stock Prediction is Artificial Neural Networks trained by Backpropagation, but this approach has a better performance when used to prediction of recent few days stock prices. But the goal of this paper is to predict monthly Maxima & Minima, so we are dealing with longer periods in which the aforementioned method does not perform satisfactory. In this paper a hybrid model explained which employs a combination of neural networks and genetic algorithms to obtain a better prediction and eliminate the problem of local minima in Neural Network.

Keywords:

Stock Prediction , Finance , Risk analysis , Genetic Algorithm , Neural Network , Backpropagation