Stock price prediction using multi-layer feed forward neural network Ramani Prakash1, Dr. Murarka P.D.2 1Associate Professor, Department of Computer Science & Engineering, Global Institute of Technology, Jaipur. Email: slrramani@rediffmail.com 2Professor, Department of Computer Science & Engineering, Arya College of Engineering and IT, Jaipur. Email: prabhu.murarka@gmail.co Online published on 22 June, 2017. Abstract As we are aware that a huge amount of capital is traded through stock market all around the world and the national economy of a country is linked with the performance of the stock market, stock market has become an investment tool not only for the strategic investors but for common people also consequently it greatly affect the everyday life of an individual. The main characteristics of all stock markets is its uncertainty i.e. one cannot predict the future value of a stock and the investor may get loss by doing some bad investment but the situation cannot be avoided. The only thing one can do is to reduce this uncertainty by using some of the stock market forecasting tool. Since price of a stock depends on several parameters which can be known or unknown, it is difficult to predict stock price [1]. Although several methods exists to predict the stock price but literature says that Artificial Neural Network is one of the technique which can predict the price of stock more accurately. To train a ANN, we use training data set which consists of inputs and outputs. The performance of the ANN depends on several parameters like number of layers in the network, number of neurons in the input layer and hidden layers, activation functions used, learning rate, momentum etc.[2]. In this paper, we have used multi-layer feed forward artificial neural network using back-propagation algorithm to predict stock price. This paper is written to create a model using Artificial Neural Networks which can predict the future prices of a stock based on past prices. Top Keywords Stock, Artificial Neural Network, backpropagation algorithm, neurons, network performance, training, validation, testing. Top |