Natural-Language-Processing

Spam SMS detection using Naive Bayes
- Cleaning text data.
- Naive Bayes implementation from scratch.
- Naive Bayes implementation using scikit-learn.
Topic Modeling - LDA
Sentiment analysis with regular classifier - GradientBoostingClassifier, GaussianNB
- Test accuracy - 72%
Sentiment analysis with LSTM
- Test accuracy - 86.5%
Attention basics
- https://arxiv.org/pdf/1706.03762.pdf
Deciphering Code with Character-Level RNN
- Pre-process text to convert encoded text to matrix of integers
- Used GRU RNN
- output - Vector of decoded text

Name		Name	Last commit message	Last commit date
Latest commit History 27 Commits
DNN-speech-recognizer		DNN-speech-recognizer
Attention Basics.ipynb		Attention Basics.ipynb
Deciphering Code with Character-Level RNN.ipynb		Deciphering Code with Character-Level RNN.ipynb
Latent_dirichlet_allocation.ipynb		Latent_dirichlet_allocation.ipynb
README.md		README.md
Spam_SMS_Classification_Naive_Bayes.ipynb		Spam_SMS_Classification_Naive_Bayes.ipynb
machine_translation.ipynb		machine_translation.ipynb
rnn_sentiment_analysis.ipynb		rnn_sentiment_analysis.ipynb
sentiment_analysis.ipynb		sentiment_analysis.ipynb

Provide feedback