shakeel ahamed shoukath ali

Mining of financial data of an organization”
Objective
The stock market is one of the most important ways for companies to raise money, along with debt markets which are generally more imposing but do not trade publicly. This allows businesses to be traded publicly, and raise additional financial capital for expansion by selling shares of ownership of the company in a public market. The prediction of stock is not easier as it has no significant rules for estimating or predicting the price of stock in the stock market. Many methods like technical analysis, fundamental analysis, time series analysis and statistical analysis etc. are all cast-off to attempt to predict the price in the stock market, but none of these methods are proved as a consistently acceptable prediction tool. The prediction models of Autoregressive feed forward Artificial Neural Networks (ANN) are applied to predict the excess return time series using lagged value. For the Artificial Neural Networks model using a Genetic Algorithm is constructed to choose the optimal topology. Artificial Neural Network(ANN) a field of Artifical Intelligence(AI) proved to be more successful way to identify unknown and hidden pattern in data which is well suited for stock market prediction. Two modules are considered, one is training session and other one is prediction based on previous trained data. In artificial neural network, back propagation algorithm is used for training session and feed forward network for predicting prices.
Conclusion
A method of predicting an organizational stock market values based on the historical financial data is proposed in this project. The purpose of this system is to predict the future stock market values. It is easy for the users to invest in stock market without any risk to their funds. The dataset were collected from the historical financial data of an organization, cause and effect relationship is generated using rapidminer tool to identify impact of variable on the target parameter (i.e. close value). Forecasting stock market financial time series data are not an easy proposition as they are found to be non-linear, volatile and fluctuating The artificial neural networks is implemented in Matlab to train and test datasets. The stock opening value and the day is taken as input and the stock closing value is obtained as network output. Linear Regression and Hebbian Algoithm are used to derive the close value of an organization. The results demonstrate that the mining of knowledge maps from company data is remarkably valuable and that tracking of the main properties of the KMs can effectively indicate crises, which is not possible by conventional risk rating methods.
Training Program