Eric Heim
Software Engineering Institute
Dr. Eric Heim is a senior machine learning research scientist at the SEI’s Emerging Technology Center. Before arriving at SEI, Heim led a basic and applied machine learning research group at the Air Force Research Laboratory, Information Directorate. Heim earned a doctoral degree in computer science in 2015 from the University of Pittsburgh. Heim’s research focuses on practical issues of applying machine learning methods to real-world environments. Such issues include: data-scarcity, model robustness, learning from structured data, and enabling models to express and reason about uncertainty in their predictions. Heim has applied his work to problems ranging from identifying man-made structures from LiDAR measurements of terrain to estimating RF signal propagation effects from aerial imagery.
Publications by Eric Heim
-
Measuring Beyond Accuracy
March 21, 2022 • Conference Paper
Violet TurriRachel DzombakEric Heim
This paper was presented at the 2022 AAAI Spring Symposium on AI Engineering.
read -
Knowing When You Don’t Know: AI Engineering in an Uncertain World
November 07, 2021 • Presentation
Eric Heim
Presents a method to improve AI system robustness that evaluates machine-learned classifiers and derives metrics that directly measure calibration performance.
read -
Knowing When You Don’t Know
November 04, 2021 • Video
Eric Heim
This short video provides an introduction to a research topic presented at the SEI Research Review 2021.
watch -
Uncertainty Quantification in Machine Learning: Measuring Confidence in Predictions
August 05, 2021 • Podcast
Eric Heim
Eric Heim, a senior machine learning research scientist at the Software Engineering Institute at Carnegie Mellon University, discusses the quantification of uncertainty in machine-learning (ML) systems.
learn more -
Robust and Secure AI
June 25, 2021 • White Paper
Hollen BarmerRachel DzombakMatt Gaston
This white paper discusses Robust and Secure AI systems: AI systems that reliably operate at expected levels of performance, even when faced with uncertainty and in the presence of danger or threat.
read -
Knowing When You Don't Know: Engineering AI Systems in an Uncertain World
December 15, 2020 • Video
Eric Heim
This presentation provides a view of new research about artificial intelligence (AI) system engineering and uncertainty.
watch -
Knowing When You Don't Know: Engineering AI Systems in an Uncertain World
November 03, 2020 • Presentation
Eric Heim
This presentation provides a view of new research about artificial intelligence (AI) system engineering and uncertainty.
read -
Poster - A Series of Unlikely Events
November 03, 2020 • Poster
Eric Heim
The poster summarizes learning from sequential behavior for activity-based intelligence and modeling human expertise.
read -
A Series of Unlikely Events: Learning Patterns by Observing Sequential Behavior (video)
November 12, 2019 • Video
Eric Heim
A Series of Unlikely Events: Learning Patterns by Observing Sequential Behavior
watch -
A Series of Unlikely Events: Learning Patterns by Observing Sequential Behavior (video)
November 11, 2019 • Video
Eric Heim
Watch SEI principal investigator Eric Heim discuss research to develop novel Inverse Reinforcement Learning (IRL) techniques as efficient and effective means for DoD/IC to perform activity-based intelligence or to teach novices how to perform tasks.
watch -
A Series of Unlikely Events: Learning from Sequential Behavior for Activity-Based Intelligence and Modeling Human Expertise
October 28, 2019 • Presentation
Eric Heim
This presentation describes work to use inverse reinforcement learning techniques to perform activity-based intelligence.
read -
A Series of Unlikely Events: Learning Patterns by Observing Sequential Behavior
October 28, 2019 • Poster
Eric Heim
This poster represents research to apply Inverse Reinforcement Learning techniques to model sequential behavior.
read -
Learning by Observing via Inverse Reinforcement Learning
March 22, 2019 • Video
Ritwik GuptaEric Heim
This SEI Cyber Talk episode explains how inverse reinforcement learning can be effective for teaching agents to perform complex tasks with many states and actions.
watch