Course Web page (this page) http://www.cs.tufts.edu/comp/150RML/
|
Instructors:
Carla Brodley:
brodley@cs.tufts.edu
and
Roni Khardon:
roni@cs.tufts.edu
Prerequisites: COMP 135 or COMP 136
Syllabus:
This course will be built around several concrete
real-world problems suitable for analysis using machine learning.
Students will be grouped into teams and will investigate/invent the
machine learning algorithms needed for the chosen application (example
applications include predictive medicine, collaboration prediction,
text classification, geography and collaborative
filtering). Background material on the applications and relevant
machine learning research literature will be provided in early
lectures. The required course work will be 1) reading current
literature related to the chosen research topic; 2) writing code with
new ideas, writing code to compare to ideas in the literature, running
experiments, understanding how to evaluate performance for the
particular application, and most importantly being creative; 3)
writing a paper on your research with your team and the instructors;
4) presenting your research to the group; and 5) one or two weekly
meetings with your group and the instructors. By taking this course a
student will obtain a more in-depth understanding of core machine
learning ideas, an appreciation for the issues that arise when
applying machine learning in practice, and experience in how to do
research in machine learning.