FINANCIAL FORECASTING USING A NEURAL NETWORK TRAINED WITH THE MACD TECHNICAL INDICATOR

摘要

This research studied the use of neural networks in financial forecasting using the MACD technical indicator. Specifically, a backpropagation neural network was trained with four inputs: closing price, volume, MACD, and signal line. Moving windows were used for each input to the network. The optimal moving window size for each input was found. Also, output threshold levels were applied to the output of the network. The results were examined. Finally, the optimal number of neurons in the hidden layer of the network was found. Output threshold levels were tested on the output of the "final" network. It was found that a neural network trained with a technical indicator can be useful in predicting the price movement of a financial market.