ABSTRACT
Since stock market prices are unpredictable, there are no consistent patterns in the data to create any near-perfect model of stock prices. Technical indicators play important roles in building a strategy, hence this research employed technical indicators, namely Accumulation Distribution Oscillator (ADOSC), Commodity Channel Index (CCI), Larry William R% (WILLR), Momentum (MOM), Rate of Change (ROC), Relative Strength Index (RSI), Simple Moving Average (SMA), Moving Average (MA), Weighted Moving Average (WMA), as the variables input in Artificial Neural Network (ANN) Backpropagation with Multivariate Regression. The models were evaluated using three statistical performance evaluation criteria. The daily data on Indonesia Stock Exchange were chosen, especially those in 10 years of trading days, to predict daily closing price. Experimental results showed that the ANN Backpropagation with Multivariate Regression obtained a promising performance in the closing price prediction on the training and validation data compared with other models.
