HW 11 (29.11) Machine Learning II, projects

1. Study the "Tennis" data set and calculate the Naive Bayes classifier for that. Would it make any mistakes on training data?

Comment from TA: "Tennis" data set is on lecture slides: Lecture 6 slide 47.

2. Think of 2-dimensional data. Generate extra features based on those (X1,X2) coordinates that would allow you to map data to higher dimensions. Make an example where hyperplane in higher dimensions would define for example circles in original 2-dimensional space. Illustrate.

3. Study the ML Gallery - http://home.comcast.net/~tom.fawcett/public_html/ML-gallery/pages/index.html Based on that find your favourite example data sets and ML methods. State some surprises to negative as well as positive direction. And conclude, how much data would be reasonable to ask and which method you would prefer to use if you were given a new data set of similar type.

4. Agree (potentially with a group) a project topic area.

Comment from TA: Write group members and project title.

5. Write a 1-page project description with clearly stated problem, data, vision, and expected goals for the project.

6. Bonus(2p) Create and agree a clear project plan, milestones, hours (at 1-2 h granularity), and deliverables for each project team member and the team as a whole. Later, during the project, keep track of hours and deadlines - declare how much you needed to deviate (and why). It would be nice to create a Google Doc for this to maintain during the project, and with a public read-only link that you could include in the final report.