Research On The Competitiveness Of Video Websites Based On Multiple Regression

Video consumption has become a way of entertainment for the general pubiic.Currently,there are many video sites on the market,and each video site is vigorously layout homemade programs,independent broadcast programs,quality programs,etc.At the same time each video site concentrates on technological innovation and user experience.Their purpose is to attact more users to use their products.In order to better understand the factors that motivate users to choose a video site,we conducted a questionnaire survey and got the data of 14 cities in China.Based on some theoretical knowledge of the multiple regression models and feature selection,the paper summarizes the key driving factor model.It should be noted that the paper is considered from the linear regression model.This model is mainly divided into two steps.The first step is to test the model's basic assumptions.By examining various hypotheses,the paper found that omission of explanatory variables can cause many problems,such as endogeneity,sequence correlation,error terms of non-normal,heteroscedasticity,etc.And the omission of explanatory variables will result in poor model fitting.Therefore,when collecting data,We should collect as many explanatory variables as possible.The second step is feature selection.Considering the explanatory variables are grouped and whether the group is known,the paper studied from three aspects.The paper gives different methods for different situations,and gives the practical guidanceFinally,we analyze the case of video competition,and draw the conclusion that the most important factor that affects the user's choice of video sites is content index,especially the content type and content coverage.The second is the brand index,mainly reflected in the user's reputation.The last is functional experience index.At the same time,we also analyze the specific performance of each video site in these indexs,and give different suggestions for different video sites.