Generation model and discriminant model

Summary:

1. Definition

2. Common algorithms

3. Differences

4. Pros and cons

　　

Content:

1. Definition

1.1 Generating the Model:

In the theory of probability statistics, the generation model is the model which can generate the observational data randomly, especially under the condition of given some implicit parameters. It assigns a joint probability distribution (joint probability distribution) to the observed and labeled data series. In machine learning, generative models can be used to model data directly, such as data sampling based on the probability density function of a variable, or to establish conditional probability distributions between variables. Conditional probability distributions can be formed by a generation model based on Bayesian criteria (reference from: Chinese wiki)

1.2 discriminant Model: Discriminant model is a method based on probability theory, which directly learns decision function y=f (X) or conditional probability distribution P (y| X) as a predictive model, i.e. a discriminant model. The basic idea is to establish discriminant function under the condition of finite sample, not to consider the model of sample generation, and to study the prediction model directly.

2. Common algorithms

2.1 Generating the Model:

Typical build models include the following:

• Mixed Gaussian models and other mixed models
• Hidden Markov model (HMM)
• Random context-independent grammars
• Naive Bayesian classifier (NB)
• Aode classifier
• Potential Dirichlet distribution model (LDA)
• Finite-Boltzmann machine

2.2 Discriminant Model

Typical discriminant models include:

• Perception Machine
• Logistic regression
• Support Vector Machine
• K Nearest Neighbor
• Decision Tree

3. Differences

Https://zh.wikipedia.org/wiki/%E5%88%A4%E5%88%AB%E6%A8%A1%E5%9E%8B

Http://dataunion.org/8743.html

http://blog.csdn.net/zouxy09/article/details/8195017

Generation model and discriminant model