Based On The Data Mining Technology In The Stock Technical Analysis Prediction System Design And Implementation

Stock market plays a very important role in China’ s economy. Stock market is avery complex system affected by the policy, economic, and investor’s psychology, andmany other complex factors. Due to its typical complex uncertainty features, the modelbuilding is very difficult. With the increasing stock data scale, the stock price implies inthese vast amounts of data. The new data processing mining technology with rapiddevelopment provides with an efficient way to obtain implicit, valuable information outfrom these masses of data, therefore the use of analysis and forecasting technology hasvery significant theoretical and practical meaning.Based on the common difficulties in domestic stock market price prediction, suchas too many technical indicators, high error rate of technical indicators commonly usedin prediction, the difficulty to judge the direction of indicator combination, thecomputational complexity of predicting process, and unsatisfying forecasting results,etc., this thesis mainly focuses on finding the integration point of data miningalgorithms and stock forecasts technical indicators through establishing predictionmodel with data mining methods in several technical indicators, and the analysis ofprediction and forecasting theory, processes and results.This thesis builds the corresponding data mining model based on features of thestock data. First, collate the abnormal data and choose certain data features as testfeature; then form the classification rule by classification calculations of the data setwith the properly adjusted decision tree classification ID3algorithm and Bayesianclassification algorithm. In the end, the classification results are to be tested in practice.The actual transaction data on the operation of the stock results show that: it is feasibleand effective using decision tree classification algorithm to predict the stock with thedata mining models. The classification and prediction algorithm of data mining aresuccessfully applied in stock technical indicators analysis, and the best integration ofclassification rule are also generated in stock price trend. With scientific application ofthis classification rule, investors are supposed to analyze technical indicators, predictstock price movements and reduce investment risk.