Evolutionary neural network for stock prediction and trading

Abstract

In recent years, various intelligent techniques, such as neural networks (NNs) and genetic algorithms (GAs) have been applied to a large variety of applications in areas of stock market prediction, trading and investment. Numerous researches have been conducted in these areas by combining various computational techniques to develop intelligent or expert systems. Nonetheless, each computational technique has its own strengths and weaknesses. For example, genetic algorithms are good for optimization, but rather poor for knowledge representation; neural networks are good for learning ability and forecasting, but lack explanatory capability. As a result, a hybrid model is needed to extract knowledge from raw data and learn to adapt to a rapidly changing investment environment.