Chapter : Lesson 4

Episode 8 - Multiclass Classifiers

face Josiah Wang

Summary:

  • How do you convert a binary classifier to handle multiclass classification?
    • One vs. all or one vs. rest: A vs (B+C), B vs (A+C), C vs (A+B)
    • One vs. one: A vs B, A vs C, B vs C
    • Reformulate model/loss function for multiclass classification
      • e.g. Categorical cross-entropy loss or Softmax loss