|
Milan Vojnovic
Professor, Chair in Data Science
|
|
Office: Columbia House, Room COL 5.05
Post: |
Department of Statistics
London School of Economics and Political Science
Houghton Street
London, WC2A 2AE
United Kingdom |
Email:
m.vojnovic@lse.ac.uk
|
Research Areas: data science, machine learning, artificial intelligence, multi-agent systems, networks
Short Bio,
Curriculum Vitae,
Publications,
Talks
Teaching:
Research interests
My current research interests lie in the broad area of machine learning with a focus on optimization and machine learning,
sequential decision making,multi-agent systems, and statistical inference in network and ranking data/
I am interested in supervising PhD studies in the research areas indicated above - if you are interested,
please don't hesitate to contact me.
My research and teaching activities have been supported by various industrial partners.
Current PhD students
- Yiliu Wang
- Jialin Yi
- Kaifang Zhou
Past visitors
- Jung-hun Kim (research intern, KAIST)
- Pierre Boudart (research intern, ENS Paris)
- Wonbong Jang (research intern, now PhD student at UCL)
- Remi Jezequel (research intern, now at INRIA)
- Konstantin Kutzkov (post-doc, now Data Scientist at Teva Pharmaceuticals)
- Zhenming Liu (visitor, Assistant Professor at The College of William & Mary)
News
- Paper on test score algorithms for utility maximization under budget constraints, with Dabeen Lee and Seyoung Yun, accepted by INFORMS Journal on Optimization, May 2022 (arxiv version)
- New working paper Sketching Stochastic Valuation Functions, with Yiliu Wang, arXiv, February 2022
- New working paper Rotting Infinitely Many-armed Bandits, with Jung-Hun Kim and Se-Young Yun, ArXiV, January 2022
- New working paper Scheduling Servers with Stochastic Bilinear Rewards, with Jung-Hun Kim, ArXiv, December 2021
- Paper Popularity Prediction for Social Media over Arbitrary Time Horizons, with Daniel Haimovich, Dima Karamshuk, Thomas Leeper, and Evgeniy Riabenko, accepted for VLDB 2022
- New paper Scheduling with Stochastic Holding Costs, with Dabeen Lee, NeurIPS 2021
- Seminar at the University of Warwick, Learning to Schedule, CS Colloquium, University of Warwick, 21 October, 2021
- New working paper Learning to Schedule, with Dabeen Lee, May 2021
- Paper Pure Exploration and Regret Minimization in Matching Bandits accepted at ICML 2021, with Flore Sentenec, Jialin Yi, Clement Calauzenes, and Vianney Perchet, May 2021
- New working paper Test Score Algorithms for Budgeted Stochastic Utility Maximization, with Dabeen Lee and Se-Young Yun, December 2020
- Facebook Systems for ML Research Award, February 7th, 2020
- Upcoming seminar at the Fourth King's Workshop on Random Graphs and Random Processes, April 3, 2020
- Upcoming seminar at the Bristol Data Science Seminar Series, University of Bristol, February 26, 2020
- Paper Communication Complexity of Approximate Maximum Matching in the Message-Passing Model, with Zengfeng Huang, Bozidar Radunovic, and Qin Zhang, accepted for Distributed Computing journal, January 22, 2020, arXiv version here
- Paper Convergence Rates of Gradient Descent and MM Algorithms for Bradley-Terry Models accepted for presentation at The 23rd International Conference on Artificial Intelligence and Statistics (AISTATS 2020), with Seyoung Yun and Kaifang Zhou
- Congrats to Sahra Ghalebikesabi for receiving the Winton Prize for Best Dissertation in MSc Statistics on the topic of scalable feature learning for dynamic networks, and the Royal Statistical Society Prize for Best Graduate (news here)!
- Congrats to Michailia Panagiotidi and Jurriaan Parie for receiving the Winton Prize for Best Dissertation in MSc Data Science on the topic of personalized learning to rank, and Michailia for receiving the Winton Prize for Academic Excellence in MSc Data Science (news here)!
- Paper A Test Score Based Approach to Stochastic Submodular Optimization with Shreyas Sekar and Seyoung Yun, accepted for Management Science, Dec 2019
- Invited seminar of the School of Science and Technology Learning to Match in Online Platforms, Athens University of Economics and Business, Athens, December 12, 2019; slides here
- Started working on a research collaboration with Facebook Core Data Science team, London, June 2019
- Congratulations to Tianlin Xu for receiving a highly commended award for her contributions as a graduate teaching assistant for our new course ST449 Artificial Intelligence and Deep Learning, May 2019
- New Data Science group created in the Department of Statistics, joining three previously existing groups, May 2019
- LSE SEDS Annual Summit 2019 to be held on May 20-21, 2019, Wolfson Theatre, LSE (event page) - featuring invited speakers, research poster sessions and other activities to learn about exciting developments in data science area !
- Workshop programme available: Workshop on Machine Learning and User Decision Making, co-organised with Laurent Massoulie, Paris, May 23-24, 2019, Paris, France
- Seminar Scalable Methods for Machine Learning Optimisation, Statslab, University of Cambridge, May 10, 2019
- Plenary talk "Optimisation for Inference and Machine Learning" at IMA and OR Society Conference on Mathematics of Operational Research, Aston University, Birmingham, April 25-26, 2019
- Check out our new SEDS website with links to people and activities, January 2019
- Our MSc Data Science capstone projects featured as an example of excellent teaching and learning practices across the School: see here
- New working paper "Convergence Rates of Gradient Descent and MM Algorithms for Generalized Bradley-Terry Models" with S. Yun and K. Zhou, Jan 2019, arxiv
- Congratulations to Kaifang Zhou ! - for winning the Winton Prize for Best Dissertation in MSc Statistics for dissertation with the title "Convergence Rates of Iterative Parameter Estimation Methods for Ranking Models"
- Keynote talk "Scalable Inference for Statistical Models of Ranking Data" at the Workshop on Challenges for Categorical Data Analysis (CCDA) 2018, RWTH Aachen University, October 22-23, 2018 ppt
- New generalized version of our working paper Maximizing Stochastic Submodular Functions by using Test Scores is now available, October 14, 2018
- Theory of ML study group @ LSE started its activities on October 2nd, 2018
- Our paper KONG: Kernels for ordered-neighborhood graphs accepted for NIPS 2018 Spotlight
- Received a 2018 Criteo Faculty Research Award, July 13th, 2018
- IEEE ICDE 2019 vice chair for the area Distributed, Parallel and P2P Data Management
- Zhenming Liu to join as a visiting scientist for six months starting in July 2018
- Lee Family PhD Scholarship awarded to Yiliu Wang from Oxford University to pursue her research in the area of machine and statistical learning
- Workshop on Data Science Theory and Practice March 27-28, 2018, Wolfson Theatre, LSE, brought together leading academic and industrial experts in data science area
- Invited talk "Scalable Machine Learning for Big Data Analytics," Huawei European Research Symposium, Paris, France, January 22-24, 2018
- Konstantin Kutzkov joined as a Research Fellow (post-doc) on January 15th, 2018 to work on machine learning for graph data representation
- Started teaching a new course lse-st446.github.io on distributed computing for big data, January 08, 2018
- Invited talk "Thurstone Choice Model Inference," Workshop on Next Generation Network Analytics, Royal Statistical Society, January 4-5, 2018
- Our new MSc in Data Science programme received its first cohort of students!
- Open post-doctoral researcher position in data science and machine learning area (now closed)
- Accepted paper "Communication-Efficient Stochastic Gradient Decent, with Applications to Neural Networks," with D. Alistarh, D. Grubic, J. Li, and R. Tomioka, NIPS 2017 (Spotlight), arxiv, featured among Microsoft @ NIPS 2017 papers
- ESRC 3+1 PhD Scholarship awarded to Kaifang Zhou on the topic statistical methods for network data, August 2017
- New working paper "Submodular Maximization using Test Scores," with S. Sekar and S. Yun, June 2017, arXiv
- New review article "Contest Theory," Communications of the ACM, Volume 60, Issue 5, May 2017, ACM Digital Library, video
- New working paper "QSGD: Communication-Optimal Stochastic Gradient Descent, with Applications to Training Neural Networks," with D. Alistarh, D. Grubic, J. Li and R. Tomioka, arxiv
- New working paper "Parameter Estimation for Thurstone Choice Models," with S. Yun, April 2017, arxiv
- New working paper "Communication Complexity of Approximate Maximum Matching in the Message Passing Model," with Z. Huang, B. Radunovic, and Q. Zhang, April 2017, arxiv
- Talk "On a Portfolio Selection Problem using Individual Test Scores," Risk & Stochastics Conference, London, April 20-21, 2017. ppt, videos
- New working paper "Adaptive Matching for Expert Systems with Uncertain Task Types," with V. Shah, L. Gulikers and L. Massoulie, March 2017, arxiv
- Talk "How to Hire a Team using Individual Test Scores?" LSE Maths Seminar on Combinatorics, Games and Optimisation, London, UK, January 12, 2017 ppt
- Blog piece Contest Theory, December 14, 2016
- Talk "Communication Efficient Stochastic Gradient Descent," Huawei workshop on algorithms for next generation networks, Paris, France, November 7, 2016 ppt
- Joined LSE, October 2016