Courses for Jan - Apr 2020

  1. Topics in Rigorous Statistical Mechanics (Elective)

    Instructor: Riddhipratim Basu

    Venue: IISc Mathematics department, room TBA

    Class Timings: Tuesday - Thursday 3:30-5:00 pm

    Course Description: This is not a physics course!! We shall cover a selection of topics in probability theory coming from statistical physics models on the Euclidean lattice. A few possible examples of the models include: Ising model, O(N) model, Gaussian free field, contact process, voter model and exclusion processes.

    Prerequisites: This course will be aimed at Int-Ph.D. and Ph.D. students working in probability theory and related areas. A course in graduate probability theory is useful, but not absolutely necessary. A student with a strong undergraduate background in probability (i.e., without measure theory) might also find this course accessible.

    For more details:http://math.iisc.ac.in/all-courses/ma397.html

  2. Learning from Data (Elective)

    Instructor:  Amit Apte and Sreekar Vadlamani

    Venue: Feynman Lecture Hall, ICTS Campus, Bangalore

    Class Timings: Mon + Wed 09:30-11:15; Lab: Tue 09:30-11:30 (Chern Lecture Hall)

    First Meeting: Wed 08 Jan 11:00-12:30

    Text Books:

    1. An Introduction to Statistical Learning, with Applications in R, by Gareth James, Daniela Witten, Trevor Hastie, Robert Tibshirani (ISLR in the rest of the document)
    2. Other references announced in class when they are needed (probably some parts of “Machine Learning: a Probabilistic Perspective,” by Kevin Murphy)

     

    Prerequisites: Basic probability theory; Linear algebra; Python (or R / Matlab / Julia, but the instructors will use and can help with python only!); Access to laptop with python / R / Matlab / Julia

    Structure of the course: The course will consist of ≈7 units of ≈7.5 hours each, consisting of lectures and labs, with approximately 2 hours of the lab for 3 hours of lectures. For each topic mentioned below, “N hours + M hours” means N hours of lectures and M of a lab.

    What should the students gain out of this course?
    On successful completion of this course, it is intended that the students would be able to perform data analytic routines involving fitting statistical models to the given dataset for different cases (qualitative and quantitative) of response and predictor variables. Specifically, the student will learn algorithms to unravel patterns in data and make predictions and inferences using data.

    How will the course achieve the goals?
    The lectures will cover the basic theory of each of the methods while the students will get hands-on experience of implementing the routines discussed in class on different datasets in lab sessions.

    What is the assessment?
    Regular homework assignments and in-class quizzes; p ≈ 40%
    Either a final exam or a project (to be decided, based on many factors): (100-p)%