Prediction of stock exchange indicators with time series and Artificial neural networks (ANN)- samples from Dow Jones industrial average, NASDAQ, STANDARD and POOR’S 500
Prediction of stock exchange indicators with time series and Artificial neural networks (ANN)- samples from Dow Jones industrial average, NASDAQ, STANDARD and POOR’S 500

Summary/Abstract: Considering the fact that the most of the stock exchange indexes are random walks it is always a big challenge to predict them. In this research are included some results of different methods for prediction based on time series analysis. Considering the amount of data accessible and the importance of the world most known indexes, as sample are taken the values of the indexes of Dow and Jones industrial, NASDAQ and Standard and Poor’s. No matter what kind of a method is used in prediction of random walks there is always risk and that risk cannot be forecasted. This is also known as random walk hypothesis. In the paper are included samples of three different indexes: Dow Jones Industrial Average, NASDAQ and Standard and Poor’s 500. The methods used for forecasting are ARIMA models and ANN or other known as artificial neural network. ARIMA models are well-known time series analysis methods and are traditionally used to forecast indexes and other time series data.

• Details
• Contents