Capstone project for Udacity Machine Learning Engineer Nanodegree
The training and testing datasets for the NLP with Disaster Tweets can be downloaded from Kaggle's competition webpage.
Note: The notebook assumes data files are stored in data directory.
- Python >= 3.6
The libraries used are:
- numpy
- pandas
- seaborn
- plotly.express
- wordcloud
- re
- nltk
- sklearn
- xgboost
It is recommended to use XGBoost with GPU support for better performance. If you need more information on that, please read here.