These are my materials for Deep Dive sessions to ML using popular Python libraries, numpy, scipy, scikit-learn, matplotlib and pandas. I summarized this from a lot of other material scattered in the web.
https://github.com/krsnewwave/DeepDive
Bonus! A very original contribution though is extracted senators’ speeches from the 2010 PH Senatorial Elections. Check it out!
https://github.com/krsnewwave/DeepDive/blob/master/data/ExtractedSenatorSpeeches.zip
You may have noticed the weeks are a little jumbled. The gaps are either for non-lecture weeks or for Kaggle sessions. The ‘stream of learning’ is the following:
- Get comfy with Ipython notebooks and numpy on week 0. Produce some viz using matplotlib.
- Go regression and simple classification on week 1. Get started with feature engineering.
- Run through of advanced models in week 2.
- Reporting in week 3.
- Advice time! Tussle with regularization in week 4. Get fancy with feature transformations. Santander Kaggle homework.
- Natural language processing on week 5. Something really awesome we did was analyze PH Senator’s speeches. Data provided!
- Random Acts of Pizza and more Santander from Kaggle for week 6.
- Unsupervised learning on week 7. Preparation for Capstone Project.
- Consultation on the Capstone project.
- An ML Competition bootcamp happened. See the MNIST folder.
- Capstone presentations!
Acknowledgements to Andrew Ng’s work, NLTK contributors, and the other sources I’ve linked to in the material.
It is the REAL DYL 