The Stock Portfolio Study Based On Association Rules And Database Dividing

In recent years, the concept of "big data" gradually comes into our vision."Big data" technology is generally called cloud technology or data mining technology. It has a very wide range of applications in finance, electricity, medical and other industries. In this paper, we begin with the traditional data mining techniques and study some association rule methods. Moreover we apply them to the stock portfolio and explain the earnings of a stock portfolio within a given period.In this paper, firstly, we introduce the Apriori algorithm and FP-Growth algorithm. Combining with the characteristic that the amount of stock data is very huge, we divide the whole dataset into some small subsets. In every subset, we mine association rules by using FP-Growth algorithm. Then we obtain the association rules relative to the entire dataset by combining the association rules relative to the subsets. When dealing with the stock data, we transform complex stock dataset to dataset which only contain0and1, since this research aims at the closing price of each stock in every trading day. Then we mine association rules by using mathematical software. At last, when we evaluate the value of rules, we give that the length of the rule has influence on the support, confidence, choice, time cost.