Data Mining Stock Investment

The China's securities market started since the last century and gradually develop and grow. At the end of 2005 the share-trading reform launched which further promoted China's securities industry to the specialization and standardization. The stock market has become a barometer of the economic situation and play an important role on the national economy and people's livelihood.As a huge amount of data processing methods, data mining and the stock market is inextricably linked. The securities industry rapidly development because of the information technology's improvement. Many businesses in stock market rely on digging out the large amount of historical data analysis to find the valuable information. Data mining can be very good to deal with these non-stationary, non-normal, high-noise data. By combining statistical sampling, estimation and hypothesis testing, machine learning search algorithms and modeling techniques, data mining have a considerable advantage in dealing with large amounts of data,high-frequency, multi-dimensional data. The most important factor is that it can continue to embody the new data to dynamic update the data model which is very suitable for the new environment.The investment in stock market related to the two points-the stock selection and time selection, the article use the classification methods of data mining and time series correlation pattern mining methods of data mining to deal with these two points respectively by empirical research. In stock selection, this article used the decision tree,neural network classification and support vector machine to classify and forecast, and comparing these different methods, we can choose out the best way to be used to predict.In time selection of the stock, this article use the way of the time series mining methods, using of the classic Apriori algorithm in association rules to mine the valuable association rules. In order to find a particular time to invest in stock market, this article use the different time series in the different time periods. This article adding time-delay threshold to consider this issue. This article lagged one day to deal with part of the time series. Through the excavation of these sequences, we can get very useful information in stock investment.This article Innovation:Taking consider of the concept of opportunity cost in the investment process and establishing the own unique method of stock assessment. This article improves the time-series association mining by taking into account the time delay threshold in the time-series mining. The means make the results more real and practical meaning. Establishing a variable system model for the investors.