Anomaly Detection Of Stock Market Based On Low-Rank Matrix Decomposition

In this paper,an anomaly detection algorithm based on the low rank Matrix decomposition principle is proposed,which is applied to a large number of data in the stock market in the financial field,this can be followed by the formulation of relevant policies,shareholders’ investment behavior and so on to make feasible recommendations.This paper puts forward the problem from theory and reality,combs the research status of outlier data mining,and analyzes the outlier data in the stock market and its causes.Firstly,the correctness of parameter selection is verified by simulation experiment.According to the principle of low rank matrix decomposition,the adaptive algorithm of anomaly detection is applied to the stock data,taking Maotai and Minsheng Bank of Guizhou as examples,this paper probes deeply into the internal information of stock data and catches the hidden anomalies therein.The experimental results prove that the abnormal information can correspond to the occurrence of major news events,it has a new practical significance to study the stable development of the stock market.That is,the anomaly detection algorithm proposed in this paper can effectively detect abnormal fluctuations in the stock market.In a word,the theory research of outlier data mining and its effective application in the stock market need to be further studied.If we want to improve the efficiency and precision of outlier detection,we need to further explore.