Financial Time Series Based On The Moving Average Algorithm

Financial time series is a set of data based on time as the object and finance as the theme, according to the old time series data of observation, research, statistical computing, it is possible to make feasible judgment for things long-term development trend, while this judgement result applying to in economic and financial research will produce certain value. But the financial time series is a burgeoning research field, people thinking from the transition of traditional way need certain time, and the financial industry itself to give people the impression is high risk and high return, so the research of financial time series problem make us highly cool psychologically and advance bravely.Financial time series is the reification of the real world things changing process. While moving average algorithm is one of the basic methods of financial time series analysis, it is unlike the other algorithms that are so complex and tedious, and our study is that it is to make the financial time series more clear to show people the location of singular points. However, financial time series can't follow the people's consciousness to change, how to accurately calculate moving average line cross points the rate of return and the hit rate, the hit rate and yield weight deflection issues, and curve fitting error calculation, are the focus of this paper.This paper from the following aspects of research:1) Moving average calculating average5and10line cross points, and finding position, and calculating the rate of return and hit rate,at last, FCM clustering to weight of the rate of return, the hit rate, close price,volume.and analyzing clustering result on the narrow and broad sense.2) On the types of definition for symmetry type, least squares curve fitting and the results of error analysis.at last, according to industry division, mapping to the macroeconomic impacting on financial markets.