Modeling And System Design Of The Stock Selection And Timing Based On The Data Mining

With the rapid development of computer and the Internet technology,more and more people are beginning to invest their assets into financial products.On the other hand,the development of the Internet brought the era of information explosion,people face huge amounts of data every day,directly resulting in low efficiency absorption and fragments of time.For this reason,many enterprises are beginning to embark on stock quantitative analysis,digging valuable information from the markets through technical methods.However,few academic institutes are working on this field.Therefore,it is significant and meaningful to combine the finance and computer science together and do some research in quantitative analysis and stock selection.In this context,some famous stock market oscillators,including the Moving Average Convergence Divergence(MACD),Relative Strength Index(RSI),Bollinger Bands(BOLL),Williams Larry %R(W %R),Stochastic(KDJ)are analyzed for the academic study and the characteristics are summarized for the system designing and implementation.In order to quantify experimental extension,a quantitative stock selection system is designed and implemented,operating on historical and nowadays stock pools according to different oscillators and periods.During the implementation,NodeJS server is made full use to utilize its eventdriven,non-blocking I/O and lightweight characteristics.Besides,due to the scale of the data,multiprocessing and algorithm improvement are also adopted in order to improve the efficiency of the system.Finally,the system designed and implemented successfully passed the correctness and completeness test by comparing with the commercial platform.The single and combined oscillators' efficiency has also been tested and 2 samples are given as examples.Meanwhile,comparison among the existing platforms is also made,which show the cross-platform system design as one of the biggest creativities.The function of academic analysis in the system is also a made-up of the existing platforms,which has a more meaningful academical value.