Instructor Interview
In the pages linked below, Prof. Rama Ramakrishnan describes various aspects of how he teaches 15.773 Hands-On Deep Learning.
- How the Course Was Created
- Building Hands-On Confidence
- The Student Projects
- Preparing Students for a World Transformed by AI
- Refining the Course from Year to Year
Assessment
Grade Breakdown
The students’ grades were based on the following activities:
- 10% Class participation
- 50% Homework assignments (2 x 25%)
- 40% Final project
Curriculum Information
Prerequisites
- Familiarity with Python at this level.
- Familiarity with fundamental machine learning concepts such as training/validation/testing, overfitting/underfitting, and regularization.
- Ideally, students should have previously taken 15.071 The Analytics Edge or 15.072 Advanced Analytics Edge or should be taking one of these courses concurrently, or have other relevant coursework or work experience.
Requirements Satisfied
15.773 can be applied toward a Master’s in Business Administration or a Master’s in Business Analytics, but is not required.
Offered
Every spring semester
Student Information
Enrollment
229 students
Breakdown by Year
Graduate students
Breakdown by Major
Almost all the students were Sloan graduate students. About two-thirds were Sloan MBA students; the remainder were from other Sloan graduate programs such as the Sloan Fellows Program, the Master of Finance program, and the Master of Business Analytics program. Since the course is regularly oversubscribed and Sloan students get priority, it has been difficult for non-Sloan students to get a seat in the class, but non-Sloan students and Harvard cross-registrants are permitted to audit the course remotely with full access to all the course materials, videos, live streams, etc.
How Student Time Was Spent
During an average week, students were expected to spend 6 hours on the course, roughly divided as follows:
In Class
Met 2 times per week for 1.5 hour per session; 26 sessions total; mandatory attendance.
Out of Class
Outside of class, students created and trained large language models to complete several homework assignments and a larger final project.
