Label classification is also called "pattern recognition". For computers, image data is the collection of data of the color and the place. So some methods are needed to get information about "line" or "human face" from the data.
By the way, the meaning of "pattern recognition" is used in a larger way. Pattern recognition is the method to judge the answer for "What is this?".
There are two steps in pattern recognition.
There are two ways to use Pattern Recognition.
For this way, we use first step of Pattern Recognition.
If we get the information of the relationship between combination of variances and pattern, it can be the big hint of Cause and Effect Analysis.
For this way, we use two steps of Pattern Recognition.
The problem to predict "A or B?" is called "Discriminant Problem".
Difference of Good Distribution by Methods
Difference of Output by Methods
NEXT Confusion MatrixTweet