The Motivation and Application of Machine Learning in Lesson 1

Machine Learning-Machine Learning

Linear Regression:
Example:
Judging the reasonable price of a house of a certain size based on the corresponding data of the current house price and size

Logistic regression:
Example:
Judging a certain patient based on the size of existing benign and malignant tumors and the age of the patient The tumor quality of

the above regression, the implicit condition is that the data must be continuous.
The above learning has a common feature, that is, learning on the basis of existing cases, which is called supervised learning.

It is mentioned that multi-dimensional machine learning with too many parameters can obtain better learning effects through dimensionality reduction.


Aggregation:
Example:
cocktail party , since there are many people speaking at the same time, how to distinguish them

Example :
Aggregate and analyze the image, obtain the image frame, and generate a 3D image according to the frame.

The above aggregation, there is no case to refer to, is called unsupervised learning


Reinforcement learning:
by defining what What is good and what is not, through multiple learning, to achieve the least punishment, the most reward, and finally get the best result.