Divergence
Last updated
Last updated
且f(x) is convex && f(1) = 0
$D_f(P||Q)$用于评估分布P和分布Q的差异。 f(1) = 0 ==> 当p(x)=q(x)时,Df = 0 f is convex ==> Df >= f(1) = 0
令$f(x) = x\log x$,则Df是KL divergence
令$f(x) = -\log x$,则Df是reverse KL divergence
令$f(x) = (x-1)^2$,则Df是Chi Square Divergence
有两个独立的分布P(x)和Q(x),KL散度用于衡量这两个分布的差异。
最小化交叉熵等价于最小化KL散度。
http://blog.sina.com.cn/s/blog_18bdda0da0102xzpw.html