Jeffrey Jackson

Dr. Jeffery Jackson earned his Ph.D. in Computer Science from Carnegie Mellon University, an MSCS from California State University Fullerton, and a BS in Mathematics and Computer Science from Oral Roberts University. With over 30 years of university teaching experience, Dr. Jackson is deeply committed to student learning and mentorship.

His research focuses on computational learning theory, bridging theoretical computer science and artificial intelligence, while also exploring related philosophical questions. He has served as the Principal Investigator on four NSF grants, funding research for numerous undergraduate and master’s students. Dr. Jackson has chaired the Computer Science Department at Duquesne University for six years previously and has recently returned to the role, demonstrating his dedication to departmental growth and excellence.

In addition to academia, Dr. Jackson has industry experience as a software engineer and manager in the aerospace and defense sector, including work at Nellis Air Force Base. He has also authored a textbook on Web technologies, which originated from his consulting work during the dot-com boom of the 1990s.

Dr. Jackson is passionate about guiding student research and senior projects, sharing selected works online with their authors' permission. He has recently begun blogging about research topics and a fun fact: Dr. Jackson holds an Erdös Number of 3!

Ph.D., Computer Science, Carnegie Mellon, 1995
M.S., Computer Science, California State University, 1981
B.S., Mathematics and Computer Science, Oral Roberts University, 1978

Probabilism holds that a rational agent's beliefs about uncertain events must conform precisely with mathematical probability theory. Rigid acceptance of probabilism implies that assumptions are necessary to justify claims of having learned correct generalizations (No Free Lunch theorems). Dr. Jackson's current research aims to show that probabilism, in its absolute form, is not a valid epistemic principle and that therefore assumption-free learning is feasible.