Financial Market Prediction
This is an incremental approach for investors seeking to identify profitable stocks using financial data.
The paper tackled the problem of predicting good stock investments by training a Self-Organizing Map neural network on over a million data points with 125 features, but no concrete prediction results or numbers were provided in the abstract.
Given financial data from popular sites like Yahoo and the London Exchange, the presented paper attempts to model and predict stocks that can be considered "good investments". Stocks are characterized by 125 features ranging from gross domestic product to EDIBTA, and are labeled by discrepancies between stock and market price returns. An artificial neural network (Self-Organizing Map) is fitted to train on more than a million data points to predict "good investments" given testing stocks from 2013 and after.