I completed my PhD in Electrical & Computer Engineering at the University of Toronto in 2017 where I was a student under Brendan Frey and part of the Machine Learning Group. I have a broad set of interests in machine learning, but my most recent research focuses on generative models of images and their applications in semi-supervised learning; neural networks that can learn sparse representations of data; and deep reinforcement learning algorithms. During my PhD, I interned for Google Brain team in 2015 where I developed the adversarial autoencoder; and Google DeepMind team in 2016 where I worked on developing deep reinforcement learning algorithms for the StarCraft II game.
I completed my Master’s at the University of Toronto in 2012 where I worked on distributed compressed sensing, supervised by Shahrokh Valaee. I received my Bachelor’s degree from Amirkabir University of Technology (Tehran Polytechnic), Iran, in 2010.
Office: BA4165, Bahen Centre for Information Technology
Email: makhzani AT psi DOT toronto DOT edu
- Oct 2017, I am joining the Vector Institute as a faculty member. [News Release]
- Sep 2017, Our paper was accepted to NIPS 2017.
- Aug 2017, Our StarCraft project and paper are released.
- Jun 2017, I gave a talk on “PixelGAN Autoencoders” at CIFAR deep learning summer school in Montreal. [video, slides]
- May 2017, New pre-print on arXiv: PixelGAN Autoencoders.