Ali Malik

Ph.D. Student at Stanford University

malikali _at_ stanford.edu

I am a PhD student in Computer Science at Stanford University (starting in 2019) focusing on Machine Learning theory and education. My research interests are in probailistic reasoning, foundations of machine learning, and building intelligent systems that help humans learn.

I completed my Bachelor's in Mathematics at Stanford University advised by Kannan Soundararajan. As an undergraduate, I did research with Stefano Ermon and Chris Piech. I also lectured a class on standard C++ programming!

I love teaching and thinking about problems that make my brain hurt. I keep a blog on the side where I write about random Maths, CS, and Philosophy related gems that come my way.

Blog

This blog is generally just my ramblings about random things. Whenever I encounter anything interesting or figure out a cool way to think about a problem, I write the results up here for personal venting and public education.

Teaching

Here are some links to the class websites of the courses I have taught at Stanford so far. If you have any questions or suggestions for improvements, please feel free to email me!

Papers and Projects

Generative Grading: Neural Approximate Parsing for Automated Student Feedback

A. Malik, M. Wu, V. Vasavada, J. Song, J. Mitchell, N. Goodman, C. Piech

Under review in Proceedings of the 33rd Conference on Neural Information Processing Systems, 2019.

Calibrated Model-Based Deep Reinforcement Learning

A. Malik, V. Kuleshov, J. Song, D. Nemer, H. Seymour, S. Ermon

Proceedings of the 36th International Conference on Machine Learning, Long Beach, USA. 2019

Using Latent Variable Models to Observe Academic Pathways

N. Gruver, A. Malik, B. Capoor, C. Piech, M. Stevens, A. Paepcke

Proceedings of the 12th International Conference on Educational Data Mining, Montréal, Canada. 2019

BlueBook: Secure, Electronic Computer Science Exams

A. Malik, B. Capoor, C. Piech

A secure, reliable, cross-platform desktop application to administer computerised examination for large classes. Has been used to administer over 5000 exams in Stanford's introductory CS classes.

DeepGIFs

A. Malik, M. Troute, B. Capoor

Exploring the effectiveness of deep generative recurrent networks in the task of understanding and generating motion. In particular, we attempt to generate GIFs of realistic mechanical motion on a synthetic dataset from an initial seed frame.