This notebook contains the most promised techniques that are implemented to classify texts ex. reviews.
- Frame the problem and look at the big picture:
- Get The Data
- Model-1 Logistic + BOW
- Model 2 NaiveBayes + BOW
- Model-3 Naive Bayes with TFIDF
- Model-4 Naive Bayes with Ngrams
- Text processing
- Model 5 cleaned text Naive Bayes BOW and
- Model 6 Fastext vector with Naive Bayes Baseline
- Model 7 - Glove vectors with Baseline model
- Model Hyperparameter tuning best models
- Model- 8 LSTM using glove word2vec
- Model 9 LSTM word2vec + early stopping
- Model 10 bidirectional LSTM
- Model 11 GRU 2 layers
- Model 12 LSTM multiple layers
- Classify text with HuggingFace pipelines
- Classify text with BERT
- Classify text with RoBERTa
- Classify text with GPT