NLP with Disaster Tweets: Summary
tech, DataScience, NLP, Summary
Project Summary
With this project, I’ll apply and practice various NLP techniques, along with trying my hands on the tidymodels framework. I will use the NLP getting started challenge on kaggle for its dataset and also evaluating model performance using submissions.
Project Goals
- Apply and improve NLP skills in R & Python.
- Learn and apply tidymodels modelling framework to do reproducible and effective science.
- Focus more on model explainability and relative comparison of different algorithms.
- Improve skills in PyTorch deep learning framework for model training and prediction.
Series
The following series of posts tackle the above project goals sequentially
Project Part 1 - NLP with Disaster Tweets: Part 1 Data Preparation
Project Part 2 - NLP with Disaster Tweets: Part 2 Nearest Neighbor Models
Project Part 3 - NLP with Disaster Tweets: Part 3 Linear Models
Project Part 4 - NLP with Disaster Tweets: Part 4 Tree-based Models
Project Part 5 - NLP with Disaster Tweets: Part 5 Deep Learning Data Preparation
Project Part 6 - NLP with Disaster Tweets: Part 6 Multi Layer Perceptron
Project Part 7 - NLP with Disaster Tweets: Part 7 Vanilla RNN and GRU