机器学习分类的一般步骤
1.get data from file
2.data preprocess
3.get data and label
4.design the model used here
5.train the model
6.validate the model
7.test the model
8.save the model with best accuracy
note: pay more attention on
1)how to deal with the source data that model can process.
In general,the size of the input of the model is [B, C, H, W],
where B is batch size,C is number of channels,H ans W is two dimension of the input data respectively.
2)how to calculate ‘in_features’ of the first fully connected layer`