Font Size: a A A

Stock Investment Value Analysis Based On PSO Clustering

Posted on:2019-03-11Degree:MasterType:Thesis
Country:ChinaCandidate:Y ZhaoFull Text:PDF
GTID:2439330599964040Subject:Financial
Abstract/Summary:PDF Full Text Request
As China's stock market continues to grow and develop,the number of listed companies has increased year by year,and the time cost for investors in China's stock market to select information for quality stocks has also increased year by year.Using cluster analysis to distinguish stock types based on financial indicators can enable investors to quickly grasp the distribution of financial characteristics in the target industry and reduce the workload of stock selection.The existing research on the clustering of listed companies for financial indicators has certain limitations.These studies have the problems of insufficient coverage of financial indicators,excessive selection of financial indicators in a certain evaluation dimension,and so on,so there is room for improvement.This paper considers the five dimensions of evaluating the operating conditions of listed companies,and allocates the number of indicators equally in each dimension.This avoids the impact of dimensional tilt on the clustering results,and constructs a financial index system that can comprehensively reflect the operating conditions of listed companies.At the same time,particle swarm clustering algorithm has attracted attention in the academic circle because of its fast convergence rate and global optimization ability.It has also been applied in medical research,biological research,agricultural research,and highdimensional data processing.Particle swarm clustering algorithm shows strong adaptability and better stability in the above research.In this dissertation,an optimization idea of particle swarm clustering algorithm,K-means Quantum-behaved Particle Swarm Optimization(KQPSO)is introduced into the research of financial indicators for the clustering of listed companies,which fills in the blank of this part.This article takes a listed company in the A-share market of the telecommunications industry as an example.KQPSO is used to cluster data based on multi-dimensional financial indicators,and gains ten types of companies with higher differentiation in terms of profitability,the quality of profitability,operational ability,solvency,and growth.These ten types of companies are analyzed in detail to provide reference for investors in China's A-share market.
Keywords/Search Tags:Stock Categorization, K-means Quantum-behaved Particle Swarm Optimization(KQPSO), Clustering analysis
PDF Full Text Request
Related items