The Quantification Model Based On Data Mining Stock Analysis Platform

With the rapid development of computer and information technology, thetechnology of information data collection and storage has gotten an unprecedentedimprovement, every fields has produced a large amount of data, how to get theinformation that we are need from them, it was difficult to solve that if just use theconventional method, but the technology of data mining can found the informationwhich have some special rules and useful, but were hidden, it is a process to discoverknowledge and get useful data, we can use this data to analyze and forecast.Stock market as an important part of the market economy, plays a very importantrole in our country. How to analysis and forecasts the stock price effectively issignificant for the financial investment. But the price is affected by many factors, soinvest stock is an uncertain and complex process. Want to establish an appropriatemodel for it has some difficulties; at same time data grows very quickly, however someregularity of the stock price just were contained. Under this situation the newlydeveloped data mining is satisfy with the problem how to find the valuable date from somany data. So use the data mining to analyze and forecast the stock is very meaningful.This thesis discusses the model of choose stock based on cash flow. Abstract analgorithm to choose stock from its quantitative theory, and build the corresponding datamining model, and preprocessing the sample data that used to chose, convert theindexes and attributes which required for construct mining, than use the algorithmtechnique of data mining and some aggregate functions in the database, combined thedatabase query to analysis and forecasts the sample date, next checked the miningresults. Last the actual results proved that use the algorithm to get from the modeltheory to choose stock it is reasonable. Customer could choose the value stock by thiskind of rules quickly, than make an advance analysis and prediction or invest directly.Finally, according to the demonstration and classification mining model algorithm todevelop a project platform that choose stock by model, and do some necessary tests offunctions and performance. It can process multidimensional data, as a tool of investorsto make invest decision. It’s a great attempt to analysis and forecast the stock date.