This is a general purpose lightweight backtest trading engine using an artificial neural network for stocks, written in modern Java 8.
This artificial neural network uses a multi layer perceptron in a format of 5 input neurons 2 hidden layers with 21 neurons each, and 1 output layer. The learning method is backpropagation. A visual of the Neural Network can be found here:
The engine uses the HTTP get class to download a CSV file, feed the values into an array, and then normalize the data into a decimal scale. From there an activation function used to transform the activation level of neuron (weighted sum of inputs) to an
output signal:
Then the trade engine determines if the output should trigger a trade. When finished iterating through the time series it takes the prediction values and time series and maps it to a chart
The Neural Network can be quite profitable however the difficulty is getting the network to predict discrepancies in future market stucture.
This is a side project and I'm not planning to extend this further.
MIT