Multi-factor Model Based On The Minimal Spanning Tree

In recent years,quantitative investment had been studied widely.Among those generalizations,multi-factor model has become an increasingly popular model,the advantage of the model is the combination of the information,we can obtain the ideal result through this.In the econophysics,the stock network was first mentioned by Mantegna,where the minimum spanning tree was used to investigate the connection of the stock.In order to study the connection of the stock,we use the minimum spanning tree to research the contruction of the stock network.In this paper,we firstly seek the choice of the investors which tend to select some factors as a reference factor by visiting the stock investors,that is to say the choice of the candidate factors.Then,we can use the multi-factor method to take out the effective factors through the effective factors of inspection.Furthermore,we employ three meathods to solve the minimum spanning tree of the stock network,including kruskal algorithm,prim algorithm,weight matrix.According to the group technology,we can model the factor of the stock,seek the solution of the model and find out the stock of the dissimilarity.