Machine Learnning Puzzles

logistic regression 标签-1和1情况下的目标，和0 1的区别?

为什么参数越小做不容易overfitting??  参数越大越容易overfitting?