Font Size: a A A

N-element Linear Logistic Discriminant And Based On The Difference Coefficient ¦Ò / ¦Ì, The Optimal Portfolio Decision-making,

Posted on:2002-02-26Degree:MasterType:Thesis
Country:ChinaCandidate:J Y ZhengFull Text:PDF
GTID:2206360032954746Subject:Operational Research and Cybernetics
Abstract/Summary:PDF Full Text Request
Fistly, the article introduces the definition of dummy variable and clarify the differences of three main methods, which usually are used to judge one sample's assortment in all kinds of departments in our country. They are mainly applied in continuous factor variables and have a high restrict of distribution type of general population.In the second chapter, we put forward the method of linear Logistic discriminant,Whose frame is Logistic probability function and core is made up of Bayesfishes discriminant and maxinum likelihood estimate. The methods can be used under both continuous and discrete type, and has little restrict on the distributional class of genemal population.In the last chapter, on the basis of Markowitz mean-variance model, We put forward the balanced theory of portfolio profit rate maximization on the condition that difference coefficient σ/μ minimize. Using Lagrange parameter method, we gain the feasible portfolio decision.Giving each example analysis in all the chapters.
Keywords/Search Tags:Categorical Variable, Logistic discriminant, difference coefficient σ/μ, maxinum likelihood estimate, balanced theory, Lagrange parameter method
PDF Full Text Request
Related items