The workshops and demos for this course are listed below along with links to materials that exist for the demo/ workshop. For objectives of and prerequisites for this class, please see the course details. Slides are available in BlackBoard.
Many resources are available in the Georgetown Analytic Machine Learning repo. As these have been compiled over time, a majority of them may not run in Python 3. You can create a Python 2 environment if you want to run through them.
All students are encouraged to complete a notebook for the examples directory of the repo. Please submit your work following the existing format (examples/github username) via a pull request.
Session 1
- Demo: SVM GUI svm_gui.py needs to be run in Python 2.7, svm_gui_py3.py is Python 3 compatible. Python 3 users running OS X 10.8 later will need to install XQuartz.
Session 2
- Demo: Tour de Scikit-learn (This notebook has been updated to be compatible with Python 3.)
- Workshop: Machine Learning from Disaster
- Workshop: Use the ML repository to build a model. Using the Wheat Classification notebook as a model, select a data set from the UCI Machine Learning Repository and work through machine learning on a data set. You are HIGHLY ENCOURAGED to work through this and submit your work as outlined above. Your ability to understand this exercise and work through it will provide valuable knowledge for your Capstone project. A data set for Classification or Regression with at least 10 attributes is recommended.
Session 3
Session 4
- Workshop: Regression techniques with energy efficiency data, Energy Efficiency Notebook and Small app to compute energy efficiency of a house
- Workshop: Clustering Flag Data
- Extra: if you would like another clustering example, see Clustering NIST headlines and descriptions