Samanta Receives NSF CAREER Award
Samanta received the award for her project, “Robustness of Inductive Reasoning Engines.” The research is multi-faceted and integrates concepts from formal methods, logic relational reasoning, and computational learning theory to develop new foundations, algorithms, and tools for the design and analysis of robust inductive reasoning engines. Her justification for the research comes from the mainstream field of machine learning (ML) and recent explosion of interest in the related area of Programming by Example (PBE). Despite the popular and prominent successes of both fields, the algorithms can pose problems due to brittleness and may lead to unexpected failures in applications. The cause of many failures in ML and PBE can be traced to the shared task of inductive reasoning. This project proposes a more principled approach to constructing inductive reasoning engines based on a formal characterization for their reliability. The problem addressed in this project focuses on robustness of these systems, specifically: is the change in the artifact learned acceptable, or at least predictable, in the presence of small changes to the set of examples?
Writer: Emily Kinsell, 765-494-0669, email@example.com, @emilykinsell