Prediction of Stock Prices Using Artificial Neural Networks
Author Name : Sneh Saini, Dr. R R Laxmi, Dr. N.P. Singh
ABSTRACT This paper presents prediction of stock prices using Artificial Neural Network (“ANN”) approach, its characteristics, classification and uses of Applications are precisely elaborated. To showcase the application of ANN, stock price of State Bank of India is taken for training and validation of data. Experimental results were obtained on stock prices using Zero-based Log-Sigmoid, Threshold, Hyperbolic Tangent, Bipolar-Sigmoid and Log-Sigmoid functions of activation functions of ANN on neuro excel predictor software based for 3 hidden layers. Finally results obtained were compared with actual data for selecting best fit on the basis of statistical tests i.e. MAE, MAPE, MSE, RMSE, RAE and RRSE which has minimum value and more votes.