Open chwangaa opened 9 years ago
Don't know how to concise explaning it, when I train with Linear model, the global correctness is 80% the first time. However, in Convnet, the confusion matrix is basically predicting everything to be 1(the first class), any help in this please?
Don't know how to concise explaning it, when I train with Linear model, the global correctness is 80% the first time. However, in Convnet, the confusion matrix is basically predicting everything to be 1(the first class), any help in this please?