Font Size: a A A

The Application Of Fuzzy C-means Clustering In The Stock Investment

Posted on:2018-06-17Degree:MasterType:Thesis
Country:ChinaCandidate:Z X SongFull Text:PDF
GTID:2348330512998360Subject:Applied Mathematics
Abstract/Summary:PDF Full Text Request
Clustering analysis is one of the most important data mining algorithms.Recently,with the great development and widespread applications of the database technology,pluralistic data appeared in the industry.One of a most important task is to classify the data properly when facing with the large-scale data.Clustering analysis is an effective algorithms to classify the data,and can also find the inner structure of data hidden in them.With the economy development,stock market has formed a certain size and the number of listed companies has also been greatly improved.In the first stage of the stock market,with stock investment speculative strongly,which cause fundamental and technical analysis loss its effectiveness,this lead to no benefit,or even serious loss of the investment.In order to choose the stocks properly,we have to evaluate the quality of the stock.Among the global world,only quantitative investment can acquire excess returns in the capital market.In this paper,we make a systematic introduction of the fuzzy clustering theory on the part of the condition,purpose and significance of the research around domestic and abroad.Moreover,we also combine theoretical research with the empirical research,then present the data structure theory hierarchically,data similarity attribute and clustering criterion function.The fuzzy clustering analysis algorithm and implementation process is the main line;firstly we study fuzzy C-mean classification data sets and C-mean value theory and fuzzy C-means clustering algorithm of fuzzy kernel clustering algorithm and fuzzy C-mean Gauss kernel object function is also solved.Then,considering the theory of investment portfolio with the single rate and explains the significance of investment combination with β,and propose β-KFCM fuzzy classification method;finally,this paper studies the basic theories of rough fuzzy C-means,and the rough fuzzy C-means clustering algorithm is also simplified.In the empirical part of this paper,six financial indicators including liquidity ratio,cash ratio,operating income growth rate,rate of return on total assets,net profit margin and asset liability ratio are chose during the 2015 fourth quarter and the third quarter of 2016,are considered during the empirical analysis to the different plate 50 stocks from A stock marke.The conclusion divided the 50 stocks into three type of investment grades,the class I deserves high quality investment value,class Ⅱ and classⅢ take the second place.
Keywords/Search Tags:Clustering analysis, Fuzzy C-means clustering, Fuzzy kernel clustering algorithm, Rough Fuzzy C-means
PDF Full Text Request
Related items