Data Analysis Project-Machine Learning Department - Carnegie Mellon University

Data Analysis Project (DAP) Requirements

The DAP enables students to learn to use machine learning for real data analysis: to solve a well-targeted research question and report the results.

The DAP consists of two parts, totaling 18 units of effort:

Part I (6 units, DAP Prep course):
I.1. Finding a research question and corresponding data set(s) to be analyzed to answer the question. The combination of question-and-data needs to be interesting, while also feasible. This step may require pragmatism given the time-limited nature of the DAP.
I.2. Based on the question and data, the DAP involves:
• representing the data to be ready to be analyzed by a computer program,
• selecting the ML algorithms/tools appropriate for the question and data.
The choice of question and data may iterate with these steps.

Part II (12 units, DAP Research):
II.1. The DAP then develops with:
• implementing the ML system,
• interpreting the output of the ML system
II.2. The DAP concludes with:
• a written report covering the question, data, representation, tools, implementation, results, and lessons learned. This report should be no more than 15 pages.
• a presentation where the student can answer questions about the project.

Organization:
1. The student works with an advisor for the DAP who is current MLD core faculty.
For PhD students, the DAP faculty advisor does not need to be, but may be, their PhD thesis advisor.

2. The final presentation of the DAP takes place at a DAP Day, and consists of a poster plus a short (~5 minute) spotlight-style oral. We will hold one DAP Day per semester, 3-4 weeks before the degree certification deadline.

3. The DAP is evaluated by the student’s DAP committee. Along with the advisor, the committee must have one more faculty member and may also have a third member. The third member can be anyone appropriate, such as an outside faculty member familiar with the data, a senior PhD student, or a postdoc working on a related topic.

4. The student receives letter grades for the DAP Prep course and the DAP Research course (the first assigned by the DAP Prep instructor, the second by the DAP committee).