machine learning - Why, in binary classification, would we map only from the input to the feature space? -
I am following slides on NLP and machine learning. On slide 7, the authors say, "In binary class fission, we can map only from input to feature space." It seems that non-binary is different from classification (previously presented in the slide), where you map from X x Y to feature space. Why do not you map with all possible combinations of X-XY for binary classification? It appears that every possible x is Y (i.e. xx -> r ^ n)
In essence - these slides are missiliding. You can treat binary classification as multi-label classification and therefore no additional restrictions may apply. However, with X-X -> F, the binary classification is simply unnecessary . As here, you give any information about classifying class 1
(there is no other option , There are only two possibilities), while there is no part of class 0
in the multi-square scenario, you do not give any real information (it is still class 2
or K
), so for some classes the features Back to Sit is a reason.
- Regardless of what is written in these slides, you can do treat binary classification as multi-category
- X Using XY -> Binary classification unnecessary
Comments
Post a Comment