hectorsalvador / predict_genre

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

predict_genre

Uses the Brown Corpus data from the nltk library to train a model that can predict the genre of a text. The classifier uses linear interpolation that combines 1-gram, 2-gram, and 3-gram models.

F1-score: 0.4039 Lambdas: 0, 0.9, 0.1 (for 1-gram, 2-gram, and 3-grams respectively)

About


Languages

Language:Jupyter Notebook 100.0%