Publications

On the Implicit Bias of Dropout.
Poorya Mianjy, Raman Arora, and Rene Vidal.
To appear in the International Conference on Machine Learning (ICML), 2018.

Streaming Principal Component Analysis in Noisy Setting.
Teodor V. Marinov*, Poorya Mianjy*, and Raman Arora. (* Equal Contribution)
To appear in the International Conference on Machine Learning (ICML), 2018.

Stochastic PCA with l2 and l1 Regularization.
Poorya Mianjy and Raman Arora.
To appear in the International Conference on Machine Learning (ICML), 2018.

Understanding Deep Neural Networks with Rectified Linear Units.
(α-β order) Raman Arora, Amitabh Basu, Poorya Mianjy, and Anirbit Mukherjee.
Proceedings of the International Conference on Learning Representations (ICLR), 2018.
[paper]

Stochastic Approximation for Canonical Correlation Analysis.
(α-β order) Raman Arora, Teodor Marinov, Poorya Mianjy, Nathan Srebro
Advances in Neural Information Processing Systems (NIPS), 2017.
[paper]

Stochastic Optimization for Multiview Representation Learning using Partial Least Squares.
Raman Arora, Poorya Mianjy, and Teodor Marinov.
Proceedings of The 33rd International Conference on Machine Learning (ICML), 2016.
[paper][code]

Teaching Experience

Fall 15: Teaching Assistant for EN600.479 Representation Learning, Johns Hopkins U.

Spring 15: Teaching Assistant for EN600.675 Statistical Machine Learning, Johns Hopkins U.

Fall 14: Teaching Assistant for EN600.475 Introduction to Machine Learning, Johns Hopkins U.

Spring 14: Teaching Assistant for CE40.725 Statistical Pattern Recognition, Sharif U. of Technology

Fall 09: Teaching Assistant for CE40.695 Stochastic Processes, Sharif U. of Technology