Stock Trend Analysis Based on Comprehensive Prediction Algorithm
Download as PDF
DOI: 10.23977/ERMSS.2019.008
Corresponding Author
Mingzhe Guo
ABSTRACT
With the rapid development of the economy, the stock market has received widespread attention from investors. Combining the advantages of k-nearest neighbor algorithm, support vector machine algorithm and time-series algorithm, this paper proposes a comprehensive prediction algorithm and applies it to the CSI 300 index's ups and downs prediction. Firstly, a neighboring algorithm and a time series algorithm are combined to obtain a rising and falling variable of the future trend. The image is converted into a two-dimensional code number by image processing, and the nearest neighbor algorithm is used to analyze whether the curve of the time series prediction is up or down; then the variable is Input into the support vector machine algorithm to get the final synthesis algorithm; finally compare with a single time series algorithm to get the final accuracy. This method can make up for the shortcomings of the original algorithm and can predict the rising and falling trend of the stock market more accurately.
KEYWORDS
Stock trend; Time series; Machine learning; Comprehensive forecast