I am a fourth year Machine Learning PhD student (now ABD) in the School of Computer Science at Carnegie Mellon University. I am co-advised by Jeff Schneider and Barnabas Poczos. I am a member of the Auton Lab and the StatML Group. Prior to CMU, I completed my B.Sc in Electronics & Telecommunications Engineering at the University of Moratuwa, Sri Lanka.

My research interests lie in the intersection of statistical and algorithmic Machine Learning. My current research spans bandit problems, Bayesian optimisation, Gaussian processes, nonparametric statistics and graphical models. As of late, I have also hopped on the deep learning bandwagon.

I am generously supported by a Facebook PhD fellowship (2017) and a CMU Presidential fellowship (2015).



Preprints

Asynchronous Parallel Bayesian Optimisation via Thompson Sampling
, , ,
[arxiv] [AutoML slides] [Moratuwa slides]

Multi-fidelity Gaussian Process Bandit Optimisation
, , , ,
[arxiv] [code] [UCL slides]

Influence Functions for Machine Learning: Nonparametric Estimators for Entropies, Divergences and Mutual Informations
, , , ,
[arxiv] [code]



Publications

Multi-fidelity Bayesian Optimisation with Continuous Approximations
, , ,
International Conference on Machine Learning (ICML) 2017   [pdf] [EP slides]

Batch Policy Gradient Methods for Improving Neural Conversation Models
, , , ,
International Conference on Learning Representations (ICLR) 2017   [pdf]

Query Efficient Posterior Estimation in Scientific Experiments via Bayesian Active Learning
, ,
Artificial Intelligence Journal (AIJ) 2017   [aij] [arxiv]
Abridged version at:   International Joint Conference on Artificial Intelligence (IJCAI) 2015   [pdf]
      IJCAI 2015 Best Paper Award (Top 2 out of 1996 submissions)   [link]

Gaussian Process Bandit Optimisation with Multi-fidelity Evaluations
, , , ,
Advances in Neural Information Processing Systems (NIPS) 2016   [pdf] [code]

The Multi-fidelity Multi-armed Bandit
, , ,
Advances in Neural Information Processing Systems (NIPS) 2016   [pdf]

Learning HMMs with Nonparametric Emissions via Spectral Decompositions of Continuous Matrices
, ,
Advances in Neural Information Processing Systems (NIPS) 2016   [pdf] [code]

Additive Approximations in High Dimensional Nonparametric Regression via the SALSA
,
International Conference on Machine Learning (ICML) 2016   [pdf] [code] [talk: video]

High Dimensional Bayesian Optimization via Restricted Projection Pursuit Models
, , ,
International Conference on Artificial Intelligence and Statistics (AISTATS) 2016   [pdf]

Nonparametric von Mises Estimators for Entropies, Divergences and Mutual Informations
, , , ,
Advances in Neural Information Processing Systems (NIPS) 2015   [pdf - coming soon] [longer version on arxiv]

High Dimensional Bayesian Optimisation and Bandits via Additive Models
, ,
International Conference on Machine Learning (ICML) 2015   [pdf] [code] [talk: video, slides]

On Estimating L22 Divergence
, , ,
International Conference on Artificial Intelligence and Statistics (AISTATS) 2015   [pdf]

Nonparametric Estimation of Renyi-Divergence and Friends
, , ,
International Conference on Machine Learning (ICML) 2014   [pdf]

Latent Beta Topographic Mapping

International Conference on Tools with Artificial Intelligence 2012   [pdf]


  Denotes joint lead authors.



Contact

GHC 8213
Machine Learning Department
School of Computer Science
Carnegie Mellon University
5000, Forbes Ave.   Pittsburgh, PA 15213

email:   kandasamy [at] cs (dot) cmu {dot} edu