Comparion Of Weighted Fusion And Sparse Partial Least Regression

With scientific and technological development, there involves many variables in the linear regression model. If a number of factors can become 0, it would have reached the purpose of variable selection. How to correctly carry out variable selection, to all fields, particularly in genetic engineering, is extremely important. In 2008, Z.John Daye and X.Jessie Jeng presented a weighted fusion. The same year, Hyonho Chun and Sunduz Keles proposed Sparse partial least squares regression(SPLS). This paper is mainly about the two kinds of variable selection methods when used to compare the correlation coefficient between variables large and small sample size, and takes a look at whether these two methods is good or not.