The University of Southern Californiaâs Institute for Creative Technologies (ICT) is a research facility, located on a creative business campus in the âSilicon Beachâ neighborhood of Playa Vista. We are world leaders in innovative training and education solutions, computer graphics, computer simulations, and immersive experiences for decision-making, cultural awareness, leadership and health.
*This position is located in Playa Vista, CA.*
We are looking for an accomplished recent PhD graduate to design explainable models for machine-learning algorithms and decision-theoretic agent models. Potential research topics include developing explanations for basic algorithms such as Bayesian networks, decision trees and clustering, and developing explainable models for advanced algorithms for reinforcement learning and POMDPs. The applications range from education, human-machine teaming, social simulation, and cyber security. The postdoctoral scholar will join a team of distinguished scientists, research programmers, Ph.D. and Masterâs students. Postdoctoral scholars will receive mentoring, support for conference travel, and the ability to work closely with graduate students. The position is for one year, with a possibility of extension for additional years.
Design explainable models for AI algorithms
Create AI models to represent joint human and computational systems.
Design and implement novel algorithms for automatically constructing and updating such models.
Design and implement novel algorithms for using the models for inference and behavior generation, as well as explaining both to non-AI experts.
Conduct experiments and analyze the data to test hypotheses and validate models.
Write and lead authorship of high-impact conference (AAMAS, AAAI, IJCAI) and journal papers (JAIR, AIJ, JAAMAS).
Support and lead graduate, undergraduate students, and summer interns in creating and testing models.
PhD (received within the last 3 years) in computer science.
Expertise in reinforcement learning, POMDPs, other quantitative reasoning methods.
Solid understanding of AI models beyond machine learning, especially decision- or game-theoretic models.
Strong programming skills in, e.g., Python, C/C++.
Previous publications on AI, especially modeling or interacting with people.
Fluent in English.
Experience in building computational (especially AI) models of human behavior is a plus.
A strong publication record in AI in top journals and conferences, in collaboration with other researchers.
Experience leading or contributing to proposals a significant plus.
Knowledge in cyber security is a plus.
Experience in design and evaluation of educational software is a plus.
The University of Southern California values diversity and is committed to equal opportunity in employment.
Serves as a research Trainee for the purpose of enhancing and developing research competencies. Participates in planning, designing and conducting highly technical and complex research projects under the direction of a supervisor. May or may not work independently. Analyzes research data and provides interpretations. Contributes to the development of research documentation for publication and/or prepares technical reports, papers and/or records
Internal Number: REQ20085436
USC is the leading private research university in Los Angeles—a global center for arts, technology and international business. With more than 47,500 students, we are located primarily in Los Angeles but also in various US and global satellite locations. As the largest private employer in Los Angeles, responsible for $8 billion annually in economic activity in the region, we offer the opportunity to work in a dynamic and diverse environment, in careers that span a broad spectrum of talents and skills across a variety of academic and professional schools and administrative units. As a USC employee and member of the Trojan Family—the faculty, staff, students, and alumni who make USC a great place to work—you will enjoy excellent benefits, including a variety of well-being programs designed to help individuals achieve work-life balance.