- Regression VS Classification
ML | ||
Regression | Classification | |
Linear Regression | Logistic Regression | |
Cost function | MSE | Cross Entropy |
Classification | Example | Function | Loss |
Linear Regression | mse | ||
Binary Classification | input : x output : yes or no |
Sigmoid | nn.BCELoss() |
Multi-class Classification | input : x or multiple x output : a or b or c... |
Softmax | nn.CrossEntropyLoss() |
ML | ||
Supervised Learning | Example | |
Training : x | Label : y(Answer) | Natural Language |
Unsupervised Learning | Example | |
Training : x | No label | LSA, LDA |
Self-Supervised Learning | Example | |
Training : x | No label-->create label | Word2Vec, BERT |
Multiple Input | ||
Multiple Linear Regression | Multiple Logistic Regression | |
activation | linear | sigmoid |
loss | mse | binary-crossentropy |
'Machine Learning' 카테고리의 다른 글
3D Tensor(Typical Natural Language Processing) (0) | 2022.08.02 |
---|---|
tensorflow (0) | 2022.03.16 |
Entropy, Cross-Entropy (0) | 2021.03.31 |
Support Vector Machine, Margin, Kernel, Regularization, Gamma (0) | 2021.03.30 |
Softmax (0) | 2021.03.24 |