Remarks on outliers in time series of stock prices based on density

Abstract

How to detect the outlier of stock prices effectively has become an intense concern of scholars in many fields. In this paper, the thoughts of Statistical theory and local density are introduced for researching the outliers in stock time series. A method detecting the outlier of stock time series based on density is proposed. Study on the outlier of time series in stock market is dealt with in two stages. In the first stage, the time series is partitioned by means of the sliding window with fixed length. In the second stage, the density-based method is introduced to detect the outliers with the local outlier factor. Empirical study, based on the daily closing price of SSE Composite Index, proved that the method is feasible and effective.