Midterm topics¶
Basic machine learning¶
Optimization: GD, SGD, convex functions
MLE
Overfitting and underfitting
F1 score
Train/dev/test split
Text representation¶
BoW representation
Feature templates
Word embeddings: skip-gram, CBOW
Language models¶
N-gram LM
Smoothing
Held-out likelihood, perplexity
Models¶
Logistic regression
Naive Bayes models
Conditional random fields
HMM (fully observered case)
Feed-forward neural nets
Recurrent neural nets
Inference¶
Viterbi decoding