Lasso与其他变量选择方法的模拟比较

Objective]To propose a new variable selection method-Lasso based on shrinkage estimate, and then compare it with other variable selection methods. [Method] First, we analyzed several common variable selection methods such as stepwise regression, AIC, BIC criteria. After stochastic simulation, we compared correct and incorrect rates of those methods.[Results] The stochastic simulation shows that Lasso's incorrect rate has decreased to 0 (n=200), with obvious advantages. As the simulation times increasing, the average correct rate of Lasso (0.951) has reached a relatively a higher level.[Conclusion] Lasso method is easy to understand, and it has broad application prospects in variable selection.