[Machine Learning]: #6 Model Selection and Cross-Validation
****************************************
Write Infront:
The topic we discuss before give us linear and polynomial functions to fit the data. But which model is the best and how can we decide which one to choose? Today we are going to tell you how to choose a better model for the data.
Let's Start Step by Step:
Here is a good video: Model Selection and Cross-Validation
*********************************************************************************
[Model Selection]:
Use the Data we have and test it by the same data, it is horrible idea!!!!
[Cross Validation]:
[Step1]: choose a model, (maybe a linear regression with d =1)
[Step2]: separate the data.
Let's Start Step by Step:
Here is a good video: Model Selection and Cross-Validation
*********************************************************************************
[Model Selection]:
Use the Data we have and test it by the same data, it is horrible idea!!!!
[Cross Validation]:
[Step1]: choose a model, (maybe a linear regression with d =1)
[Step2]: separate the data.
[Step3]: For every data set they have to once test set. and get the error value(it is not official, but you know what I mean)
................ (after all these steps, calculate the average value of you mode1)
[Step4]: Choose differently model(maybe d=2,3....n) and get diferent average value (Cross Validation value)
[Step5]: Select the model with min Cross-validation value
[Can We Modify?]:
Ans: Sperate The Data at first step, part1 is for model selection and part2 is for test
how to select the model? previous method!!!!
Comments
Post a Comment