本节主要的内容是描述琴生不等式(Jensen’s Inequality)。有关琴生不等式的描述为,如果函数$f(x)$为凸函数(convex function),那么有$\mathbb{E}[f(x)]\geq f(\mathbb{E}[x])$。
Jensen’s Inequality中的证明}
\[\begin{figure}[H] \centering \includegraphics[width=.5\textwidth]{微信图片_20191021084621.png} \caption{函数和它在某点的切线的表达式} \label{fig:my_label_1} \end{figure}\]设切点的横坐标为$\mathbb{E}[x]$,那么$f(\mathbb{E}[x])=L(x)=a\mathbb{E}[x]+b$。又因为function为convex function。那么很显然,对于$\forall x$都有$f(x)\geq L(x)$。然后,我们同时对不等式两边求期望,可以得到$\mathbb{E}[f(x)]\geq \mathbb{E}[L(x)]$。那么我们进行如下的推导:
\(\begin{equation} \begin{split} \mathbb{E}[f(x)]\geq & \mathbb{E}[L(x)] \\ = & \mathbb{E}[a\mathbb{E}[x]+b] \\ = & a\mathbb{E}[x] + b \\ = & f(\mathbb{E}[x]) \\ \end{split} \end{equation}\) 可以很简单的得出结论。
Jensen’s Inequality的推广}
假设$c$是$[a,b]$之间的任意一点,我们可以很自然的描述为$c=ta+(1-t)b$。我们可以使用Jensen’s Inequality的推广形式。这个不等式的证明很简单,结论也很直觉性(肯定是这样的),此处不再做更多的描述。Jensen’s Inequality将在后续相关的知识中大量的使用,故此处暂做补充。
\[\begin{equation} tf(a)+(1-t)f(b)\geq f(ta+(1-t)b) \end{equation}\]FEATURED TAGS
Linear
Regression
Math
Basis
Exponential
Family
Distribution
Classification
Bayes
Background
Inference
Conclusion
Feedforward
Neural
Network
Support
Vector
Machine
Margin
and
Solution
Weak
Duality
Kernel
Method
of
Function
Probability
Graph
Markov
Variational
Algorithm
Belief
Process
Introduction
Gaussian
Expectation
Maximization
Mixture
Model
Chain
Monte
Carlo
Sampling
Hidden
Learning
Kalman
Filter
Conditional
Random
Field
Boltzmann
Deep
Generative
Adversarial
Reinforcement
Blog
docker
cuda