Alchemy is a C++ application for predicting movement of securities in the public stock market. It includes the following features:
- Neural network implementation as multi-layer perceptron with back propagation
- Yahoo Finance data connector for retrieving historical data
- Usage of common technical analysis indicators as NNet inputs (e.g. moving average, stop-and-reversal)
- Creation and optimization of predictor profiles using historical data
- Graphical plotting of predictor results vs. actuals using GNUPlot
- Portfolio construction and optimization using downside risk metrics
The application is in an alpha state in that it is functionali but the interface is rough and usage documentation is incomplete. It is provided as an example of my C++ coding work.
Copyright (c) Chuck Smith 2015