An in depth analysis of neural network with application in finance

Abstract

Arti cial neural network model is inspired from how the nervous system of brain works. If it is designed properly it can process large amount of data or information and can give proper output for di erent application like pattern recognition, forecasting disease or nancial data etc. In recent years Arti cial neural network has been a great choice to analyze nancial time series data as they are quite capable of learning the relationships among di erent features of data. As the world's economy is continuously changing, there is a need for keep an eye on the dynamic conditions of economy. Therefore, nancial institutions and investors always wants a reliable system to monitor the data relationship so that they can simulate and predict - nancial positions on the basis of market trends in order to nd where should they invest. But because of the high volatility and high non linearity it has been quite a challenge to predict the nancial stock market. In this paper we attempt to study neural network and how they are actually useful in predicting stock market and nally we are going to use di erent model of ANN to predict the stock price of Amazon and SP 500 index. We will analyze the capability of neural net to cope with the nonlinear and chaotic patterns of data and their ability to predict.