Article directory
Hi, hello. I am Cha Ling.
Today let us learn about "Bayesian Statistics" and "Machine Learning Classification Indicators". After these two parts, our "Probability and Statistics" part is over. I don’t know if the content during this period is helpful to everyone.
Okay, let's get started.
Bayesian Statistics
Bayes is a very powerful person. Bayes has made too many contributions in statistics. Including probability, there are frequentists and Bayesians. Of course, there are also Bayesians in machine learning. It is to consider machine learning models from a Bayesian perspective, which is in opposition to the connectionism of neural networks.
When it comes to Bayesian statistics, there is no way around Bayesian formula, a formula proposed by a great mathematician and statistician. The formula named after him is as follows:
P(Y|X) = P(X|Y)P(Y) / P(X)
This formula is very simple and is obtained from the formulas of conditional probability and joint probability. Why can we get this formula? Let's multiply the denominator P(X) to the left side of the equation, and multiply the left side P(Y|X) by P(X). Isn't it the joint probability of X and Y? The left and right sides of the equation are a joint probability, so take P(X) on the right side as the denominator, and this formula comes out. It's that simple, Bayes' formula is that simple.
Let’s take a look at some of the meanings of each quantity in this formula. According to Bayesian language, X represents an observable sample. The observed sample can be seen and counted&#x