梯度下降算法过程以及感知机算法与梯度下降算法区别
一、梯度下降算法过程:
1)Start withy random weights: w1, w2, ..., wn, b
2)For every point (x1, x2, ..., xn):
i) for i = 1, ..., n
update wi'
update b'
3)Repeat until error is small
二、感知机算法与梯度下降算法区别
1、感知机算法只对分类错误的点才会更改权重
2、梯度下降算法对所有点更新权重