Homework for Statistical NLP class of Charles University.
- Q1: Entropy of a text (cz vs. en)
- Q2: Cross-Entropy and Language Modeling (Estimation Maximization algorithm, Linear Interpolation Smoothing)
- Q1: Best Friends (cz & en) (finding best association pairs with PMI)
- Q2: Word and Tag Classes (cz & en) (Hierarchy of words and tags using MI, algorithm from Brown et al. paper)
- Q1: Brill's Tagger & Tagger Evaluation (nltk Brill`s tagger implementation)
- Q2: Unsupervised Learning: HMM Tagging (supervised and unsupervised HMM tagging, Baum-Welch training, Viterbi decoding)