Basic Information
- Course Code: SC 654
- Course Name: Social learning and herding
- Course Offered In: 2023-2024
- Semester Season: Spring
- Instructors: Prof. Ankur Kulkarni
- Prerequisites: Familiarity with Probability theory / EE325
- Difficulty (1 being easy and 5 being tough): 4
Course Content
Basic Bayesian learning and inference framework: Introduction to Bayesian learning with binary and Gaussian models, and the martingale convergence theorem.
Social learning with common memory: Basic model of learning from others’ actions, including how individual actions communicate information and the impact of noise on social learning.
Cascades and herds: Study of how agents observe others and make sequential decisions, leading to cascades where no learning occurs, and herds where all agents eventually make the same decision.
Outcomes: Examination of social learning through observing outcomes of others’ actions and the conditions under which beliefs converge to the truth.
Sequences of financial trades: Analysis of how agents’ trades reveal private information about asset values.
Gaussian financial markets: Investigation of asset demand in CARA-Gauss settings, including models of limit orders and market orders, and the impact of social learning on trading behavior.
Feedback on Lectures
The lectures were good since it was a small class; hence, a lot of interaction with the professor made the math in the course less daunting than it was supposed to be. No tutorials. It was hard to keep up with the small changes that occurred with the models being studied. The doubts raised in class were interesting and helped me understand the subject more. The professor made sure to recap a little from the previous lecture to maintain continuity.
Feedback on Evaluations
The grading scheme for the course consisted of two homework assignments, one before the mid-semester and one after, which together contributed to 25% of the total grade. There was no mid-semester examination. The project, conducted in groups of up to three students, accounted for another 25% of the grade and involved reading and surveying papers within the subject area. The remaining 50% of the grade was based on the end-of-semester examination. Marks for individual components were not provided, and final grades were directly assigned.
Study Material and Resources
The class basically revolved around this book called “Rational Herds: Economic Models of Social Learning” by Christophe Chamley. It always compared new topics with stuff we already learned, which made it easier to grasp and connect the dots between different ideas.
Follow-up Courses
This was the first offering of the course. You can take a look at Game theory related courses as well.
Final Takeaway
Good course, a little math heavy. Going to classes can make following the math easier.