K.G. Srinivasa, K. Sridharan, P. Deepa Shenoy, K.R. Venugopal, and L.M. Patnaik (India)
Machine learning, Kohonen network, Genetic algorithms, Data mining, Prediction, Clustering.
Stock market prediction is a complex and tedious task that involves the processing of large amounts of data, that are stored in ever growing databases. The vacillating nature of the stock market requires the use of data mining tech niques like clustering for stock market analysis and predic tion. Genetic algorithms and neural networks have the abil ity to handle complex data and are immune to noise in the input. In this paper, we propose an algorithm Evolution ary Approach to Self Organizing Map(EASOM) to cluster stock market data. Genetic algorithms are used to train the Kohonen network for better and effective prediction. Fuzzy logic is used to fix learning rate for faster convergence. The algorithm was tested on real stock market data of compa nies like Intel, General Motors, Infosys, Wipro, Microsoft, IBM, etc. The algorithm consistently outperformed regres sion model, backpropagation algorithm and Kohonen net work in predicting the stock market values.
Important Links:
Go Back
