sesiria的博客

研究机器学习的时候，在markdown中最常的应用是写数学公式，和证明。一， 数学证明块：$$\begin{array} {l}{\text { Recall the exponential family form of the Bernoulli distribution }(6.113 \mathrm{d}),} \\ {\qquad p(x | \mu)=\e...

生成模型生成模型：在概率统计理论中， 生成模型是指能够随机生成观测数据的模型，尤其是在给定某些隐含参数的条件下。它给观测值和标注数据序列指定一个联合概率分布。在机器学习中，生成模型可以用来直接对数据建模（例如根据某个变量的概率密度函数进行数据采样），也可以用来建立变量间的条件概率分布。条件概率分布可以由生成模型根据贝叶斯定理形成简而言之：模型可以生成数据。基于GAN(Generat...

Reference: mathematics for machine learningHomomorphism:Injective, Surjective, Bijective.Concept for intuition:For the Identity mapping:In mathematics, an identity function, also ca...

Let's consider the example of finding the minimum of the function,along the curve (or, subject to the constraint),The functions themselves are fairly simple, on a contour map they look a...

由于高数是十多年前学的，最近在看许多NLP和ML相关的文章遇到了一些微积分卡壳了。这里找一份资料作为参考。主要是单变量微积分。多变量微积分的参考资料会在后续进行更新。这是宾夕法尼亚大学大学本科高数的参考wiki之一。当然初等微积分（微分，积分）可能还需要参考其他资料http://calculus.seas.upenn.edu/由于服务器后台更新， 无法正常浏览数学公式。可能需要安...

Chapter 6 COUNTING AND PROBABILITY6.1 IntroductionRandom Process, Sample Space, Event And ProbabilityRandom Process, when it takes place, one outcome from some set of outcomes is sure to occur, but it...

转载自：在神经网络里经常使用sigmoid做激活函数，它的导数是怎么样求解呢？因为要使用它的导数来计算梯度下降。这个过程如下：1. sigmoid函数：f(z) = 1 / (1 + exp( − z))导数：f(z)' = f(z)(1 − f(z))求导过程如下：关于sigmoid函数在神经网络中的应用：http://blog.youkuaiyun.com/zhishengqianjun/article/

Chapter5 SET THEORY5.1 Basic Definitions of Set TheoryThe axiom of extension says that a set is completely determined by its elements; the order in which the elements are listed is irrevle

Chapter 4 SEQUENCES AND MATHEMATICAL INDUCTIONOne of the most important task of mathematics is to discover and characterize regular patterns, such as those associated with processes that are repeate

转载自 http://blog.youkuaiyun.com/discreeter/article/details/69833579裴蜀定理： 对任何a,b∈Z和它们的最大公约数d，关于未知数x和y的线性不定方程（称为裴蜀等式）：ax+by=c有整数解(x,y)当且仅当d∣c，可知有无穷多解。特别地，一定存在整数x,y，使ax+by=d成立。 推论： a,b互质的充要条件是存在

Chapter3 Elementary Number Theory and Methods of proofThe underlying of this chapter is mainly about the question of how to determine the truth or falsity of a mathematical statement.3.1 Dir

The Logic Of Quantified Statementspredicate calculusstatement calculus(propositional calculus)2.1 Introduction to Predicates and Quantified Statements IForm

Chapter1 The logic of compound statement1.1 Logic Form and Logic Equivalence.Translate the natural english argument to logic notation.such as if p or q, then r.Def1:Proposition(state

1. 证明log X 0都成立采用了数学归纳法 1） 当  0 由于 log 1 = 0 ， 当X 所以log X 2） 当 x > 1时 采用数学归纳法（假设 当 x > 1时， log x log(2x) = log2 + logX = 1 + logX 由于当log x ，x >1时。他是一个递增函数 所以对于任意的 A 1 且