About me

I am a Ph.D. student and a CMU Presidential Fellow (2018) from the School of Computer Science, Carnegie Mellon University. Currently, I am working with Dr. Jaime Carbonell on developing efficient algorithms for Multitask and Lifelong learning (aka Continual learning). My current research interests are in Online learning, Multi-armed bandits, (Multi-agent) Reinforcement learning, Hyperparameter optimization, etc with applications to Natural Language Understanding.

I am in the job market now! Here is my CV and my research statement.

You can learn more about my current research from my thesis proposal document titled, Online and Adaptive Methods for Multitask Learning. My thesis committee includes an awesome group of researchers working in areas closely related to my current research: Avrim Blum (TTI-Chicago), Jaime Carbonell (Chair), Louis-Philippe Morency and Barnabás Póczos.

Preprints

K. Murugesan; J. Carbonell, Lifelong Learning with Output Kernels, arxiv.

K. Murugesan; J. Carbonell; Y. Yang, Co-Clustering for Multitask Learning, arxiv. (code)

Recent Publications

K. Murugesan; J. Carbonell, Active Learning from Peers, 31th Annual Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, California, 2017. (code)

M. Kshirsagar; K. Murugesan; J. Carbonell; J. Klein-Seetharaman, Multitask matrix completion for learning protein interactions across diseases , Journal of Computational Biology. January 2017, ahead of print. (code) (Journal version of RECOMB 2016)

K. Murugesan; J. Carbonell, Self-Paced Multitask Learning with Shared Knowledge, 26th International Joint Conference on Artificial Intelligence (IJCAI-17), Melbourne, Australia, 2017. (code)

K. Murugesan; J. Carbonell, Multitask Multiple Kernel Relationship Learning, 17th SIAM International Conference on Data Mining (SDM 2017), Houston, Texas, USA, 2017. (code)

K. Murugesan*; H. Liu*; J. Carbonell; Y. Yang, Adaptive Smooth Online Multitask Learning, 30th Annual Conference on Neural Information Processing Systems (NIPS 2016), Barcelona, Spain, 2016.

M. Kshirsagar; J. Carbonell; J. Klein-Seetharaman; K. Murugesan, Multitask matrix completion for learning protein interactions across diseases , 20th Annual International Conference on Research in Computational Molecular Biology (RECOMB 2016), Los Angeles, USA, 2016. (code)

K. Murugesan; J. Carbonell, Predicting Workplace Incidents with Temporal Graph-guided Fused Lasso, CMU LTI Technical Report 2015.

Copyright © 2018 Keerthiram Murugesan