Introduction, Set theory, Measure theory, Probability, Random variable

2021. 5. 26. 16:25베이지안 딥러닝

edwith.org

PDF download

book link

 

1. Introduction

probability, random variable, random process, kernel function에 대해서 알아보자

 

2. Set

set, element, subset, universal set, set operations

disjoint

partition

Cartesian product

power set

cardinality |A|: finite, infinite, countable, uncountable, denumerable (countably infinite)

자연수, 실수는 countable, [0,1]사이의 실수집합은 uncountable

mapping, domain, co-domain, image, range, inverse image

one-to-one=injective, onto=surjective, invertible

 

3. Measure theory

σ-algebra = σ-field 정의
properties 특징

Sigma field가 없으면 measure를 정의할 수 없다.

measure 정의

measurable space : $(U,B)$

measure space : $(U, B, \mu)$

 

4. Probability

Probability measure : measure + $\mu(X) = 1$

Bayes' rule
Posterior, Prior probability

independent != disjoint, mutually exclusive

 

5. Random Variable

Random Variable 정의

Probability density function : p.d.f

Conditional expectation, probability $X|Y$

Expectation의 정의

두 분포가 같다는 것은 mean값만 가지고 판단 할 수 없다. n-th momentum을 보고나서 판단한다.(2nd momentum: variance)

Independent => uncorrelated

uncorrelated =\=> independent

 

'베이지안 딥러닝' 카테고리의 다른 글

Gaussian process latent variable model(GPLVM)  (0) 2021.06.07
Gaussian Process의 Weight space view | Function space view  (0) 2021.06.03
Functional analysis  (2) 2021.05.28
Random process  (1) 2021.05.27
donaricano-btn