deep learning概率论基本概念

1. Aprobability distributionis a description of how likely a random variable orset of random variables is to take on each of its possible states. The way wedescribe probability distributions depends on whether the variables are discrete orcontinuous.
2. A probability distribution over discrete variables may be described using aproba-bility mass function(PMF).
3. When working with continuous random variables, we describe probability distri-butions using a probability density function (PDF) rather than a probabilitymass function.
4. Marginal Probability
   It is the probability of the occurrence of single event , it can be considered as an unconditional probability. It is not conditioned on another event.This probability is also calculated using the contingency table(also called joined probability table). Again consider the table :-
   Image for post
   From the above table , we can see four possible single event happening which can be considered as a single