DisasterClf

Description

These code are used to train the model used for the competition "Real or Not? NLP with Disaster Tweets"

https://www.kaggle.com/c/nlp-getting-started/overview/description

Approach

1. Normal Machine Learning
2. Topic Modeling output as features to train Supervised Machine Learning
3. Neural Network (GloVe, LSTM)

Accuracy

1. Bernoulli Naive Bayes with Lemmatization, OneHotEncoder, CountVectorizer -> 78.92 %
2. 222 Topic Unigram with Random Forest Classifier -> 61.87 %
3. GloVe With LSTM -> 79.20 %

Conclusion

By far the approach that generate the best accuracy is Neural Network using GloVe with LSTM