Plain Stock Prediction

Plain Stock Price Prediction via RNNs with Graves LSTM unit.

Training and Predicting a specific feature by setting PriceCategory in com.isaac.stock.predict.StockPricePrediction.java as:

PriceCategorycategory = PriceCategory.CLOSE; // CLOSE: train and predict close price// orPriceCategorycategory = PriceCategory.OPEN; // OPEN: train and predict open price// ...

The PriceCategory enum:

publicenumPriceCategory {
OPEN, CLOSE, LOW, HIGH, VOLUME, ALL
}

Predicting all features as:

PriceCategorycategory = PriceCategory.ALL; // ALL: train and predict all features

Demo Result

A Useful GitHub Repository: timestocome/Test-stock-prediction-algorithms, which contains much information, methods and sources about predict stock and market movements.