Amrita Gupta

About

Amrita is a researcher and data scientist at the Microsoft AI for Good Research Lab (opens in new tab).  Her work centers on co-developing machine learning and computational models with decision-makers and practitioners in biodiversity conservation and sustainability.

Amrita received a Ph.D. in Computational Science and Engineering (opens in new tab) from the Georgia Institute of Technology (opens in new tab), where she was advised by Professor Bistra Dilkina (opens in new tab). She then worked as a Data Scientist in the Analytics Lab (opens in new tab) at Conservation Science Partners (opens in new tab), a non-profit research organization established to meet the analytical and research needs of diverse stakeholders in conservation projects. From a technical standpoint, Amrita is especially interested in methods that facilitate the effective use of limited data and resources, such as: transfer learning to train models with limited labeled data; active learning to identify the most informative data samples for training models; out-of-distribution generalization to maximize the utility of models trained on localized datasets; and mathematical optimization to help managers and practitioners make systematic and informed decisions about the allocation of resources.