Scholar
Tim Kaler
Google Scholar ID: Ps_pNbYAAAAJ
Research scientist in computer science, MIT CSAIL
Citations & Impact
Citations
2,010
H-index
10
i10-index
10
Publications
20
Co-authors
0
Contact
Resume (English only)
Academic Achievements
- Published 'Accelerating training and inference of graph neural networks with fast sampling and pipelining' in Proceedings of Machine Learning and Systems 4, 2022.
- Published 'Evolvegcn: Evolving graph convolutional networks for dynamic graphs' in Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 34, 2020.
- Published 'Scalable Graph Learning for Anti-Money Laundering: A First Look' as an ArXiv preprint, 2018.
- Published 'PARAD: A Work-Efficient Parallel Algorithm for Reverse-Mode Automatic Differentiation' in Symposium on Algorithmic Principles of Computer Systems, SIAM, 2021.
- Published 'Executing dynamic data-graph computations deterministically using chromatic scheduling' in ACM Transactions on Parallel Computing (TOPC) 3(1), 2016.
- Published 'Polylogarithmic Fully Retroactive Priority Queues via Hierarchical Checkpointing' in Workshop on Algorithms and Data Structures, 2015.
Research Experience
- Currently leads the xGraph systems team together with Alexandros-Stavros Iliopoulos and Jie Chen with the MIT-IBM Watson AI Lab, and is a contributor to the OpenCilk project. His current research focuses on the design of fast machine learning systems for operating on million and billion-scale graphs.
Education
- Completed his PhD thesis in 2020 on 'Programming technologies for engineering quality multicore software' at MIT.
Background
- Postdoctoral associate at MIT in the EECS department and a member of the Supertech research group within the Computer Science and Artificial Intelligence Laboratory (CSAIL). His research aims to make it easy to develop fast parallel systems that have simple semantics, provable performance guarantees, and good real-world performance. He enjoys finding simple ways to express performance-engineering techniques so that they are easy to understand and apply by non-expert programmers.
Co-authors