The AWS Machine Learning Research Awards (MLRA) aims to advance machine learning (ML) by funding innovative research and open-source projects, training students, and providing researchers with access to the latest technology. Since 2017, MLRA has supported over 180 research projects from 73 schools and research institutes in 13 countries, with topics such as ML algorithms, computer vision, natural language processing, medical research, neuroscience, social science, physics, and robotics.

On February 18, 2020, we announced the winners of MLRA’s 2019 Q2/Q3 call-for-proposal cycles. We’re now pleased to announce 28 new recipients of MLRA’s 2019 Q4 call-for-proposal cycle. The MLRA recipients represent 26 universities in six countries. The funded projects aim to develop open-source tools and research that benefit the ML community at large, or create impactful research using AWS ML solutions, such as Amazon SageMaker, AWS AI Services, and Apache MXNet on AWS. The following are the 2019 Q4 award recipients:

RecipientUniversityResearch Title
Anasse BariNew York UniversityPredicting the 2020 Elections Using Big Data, Analyzing What the World Wants Using Twitter and Teaching Next Generation of Thinkers How to Apply AI for Social Good
Andrew Gordon WilsonNew York UniversityScalable Numerical Methods and Probabilistic Deep Learning with Applications to AutoML
Bo LiUniversity of Illinois at Urbana-ChampaignTrustworthy Machine Learning as Services via Robust AutoML and Knowledge Enhanced Logic Inference
Dawn SongUniversity of California, BerkeleyProtecting the Public Against AI-Generated Fakes
Dimosthenis KaratzasUniversitat Autónoma de BarcelonaDocument Visual Question Answer (DocVQA) for Large-Scale Document Collections
Dit-Yan YeungHong Kong University of Science and TechnologyTemporally Misaligned Spatiotemporal Sequence Modeling
Lantao LiuIndiana University BloomingtonEnvironment-Adaptive Sensing and Modeling using Autonomous Robots
Leonidas GuibasStanford UniversityLearning Canonical Spaces for Object-Centric 3D Perception
Maryam RahnemoonfarUniversity of Maryland, BaltimoreCombining Model-Based and Data Driven Approaches to Study Climate Change via Amazon SageMaker
Mi ZhangMichigan State UniversityDA-NAS: An AutoML Framework for Joint Data Augmentation and Neural Architecture Search
Michael P. KellyWashington UniversityWeb-Based Machine Learning for Surgeon Benchmarking in Pediatric Spine Surgery
Ming ZhaoArizona State UniversityEnabling Deep Learning across Edge Devices and Cloud Resources
Nianwen XueBrandeis UniversityAMR2KB: Construct a High-Quality Knowledge by Parsing Meaning Representations
Nicholas ChiaMayo ClinicMassively-Scaled Inverse Reinforcement Learning Approach for Reconstructing the Mutational History of Colorectal Cancer
Oswald LanzFondazione Bruno KesslerStructured Representation Learning for Video Recognition and Question Answering
Pierre GentineColumbia UniversityLearning Fires
Pratik ChaudhariUniversity of PennsylvaniaOffline and Off-Policy Reinforcement Learning
Pulkit AgrawalMassachusetts Institute of TechnologyCuriosity Baselines for the Reinforcement Learning Community
Quanquan GuUniversity of California, Los AngelesTowards Provably Efficient Deep Reinforcement Learning
Shayok ChakrabortyFlorida State UniversityActive Learning with Imperfect Oracles
Soheil FeiziUniversity of Maryland, College ParkExplainable Deep Learning: Accuracy, Robustness and Fairness
Spyros MakradakisUniversity of NicosiaClustered Ensemble of Specialist Amazon GluonTS Models for Time Series Forecasting
Xin JinJohns Hopkins UniversityMaking Sense of Network Performance for Distributed Machine Learning
Xuan (Sharon) DiColumbia UniversityMulti-Autonomous Vehicle Driving Policy Learning for Efficient and Safe Traffic
Yi YangUniversity of Technology SydneyEfficient Video Analysis with Limited Supervision
Yun Raymond FuNortheastern UniversityGenerative Feature Transformation for Multi-Viewed Domain Adaptation
Zhangyang (Atlas) WangTexas A&M UniversityMobile-Captured Wound Image Analysis and Dynamic Modeling for Post-Discharge Monitoring of Surgical Site Infection
Zhi-Li ZhangUniversity of MinnesotaUniversal Graph Embedding Neural Networks for Learning Graph-Structured Data

Congratulations to all MLRA recipients! We look forward to supporting your research.

For more information about MLRA, see AWS Machine Learning Research Awards or send an email to [email protected].


About the Author

seo yeon shin 100Seo Yeon Shin is a program manager for the AWS AI Academic Programs.