Research And Application Of Similarity-based Mining Of Financial Data Analysis System

With the deepening of economic reform and the development of China's financial market, financial products are becoming more and more popular with ordinary people. Different financial systems such as stocks, bonds, foreign exchange and futures will generate a large number of online transactions. The analysis of these transactions is already a hot issue in computer technology.Existing financial data software mainly focuses on the display of the data rather than the analysis of it. This article applies the data mining technique into the data analysis system and takes stock data for example. It mines historical data for users'reference. The article can be roughly divided into two parts. First it focuses on the key problems in data analysis which are financial data representation and the similarity measure. It compares the existing representation and similarity measure methods of time series data, proposes a suitable method for financial data, and verifies it through experiments. Next, it designs the architecture of the financial data analysis system based on software engineering concepts, making it structured, extensible and reusable. And then applies those methods which are discussed in the first part into the system as well as discusses the difficulties and technology in the implementation of the system.