Brendan Juba
Scholar

Brendan Juba

Google Scholar ID: 5wppdUoAAAAJ
Associate Professor, Washington University in St. Louis
Theoretical Computer ScienceArtificial Intelligence
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
1,212
 
H-index
17
 
i10-index
23
 
Publications
20
 
Co-authors
0
 
Publications
7 items
Near-Optimal Cryptographic Hardness of Learning With Homogeneous Halfspaces Under Gaussian Marginals
2026
Cited
0
Lifted Relational Probabilistic Inference via Implicit Learning
2026
Cited
0
The Limitations and Power of NP-Oracle-Based Functional Synthesis Techniques
2025
Cited
0
Personalized Prediction By Learning Halfspace Reference Classes Under Well-Behaved Distribution
2025
Cited
0
Polynomial-Time Relational Probabilistic Inference in Open Universes
2025
Cited
0
Distribution-Specific Agnostic Conditional Classification With Halfspaces
2025
Cited
0
Enhancing Numeric-SAM for Learning with Few Observations
arXiv.org · 2023
Cited
0
Resume
Background
  • Associate Professor at the School of Engineering, Washington University in St. Louis, in the Department of Computer Science and Engineering and the Division of Computational & Data Sciences
  • Affiliated with the Theoretical Computer Science and Machine Learning & Artificial Intelligence research groups
  • Research focuses on theoretical approaches to artificial intelligence grounded in algorithmic theory and computational complexity
  • Has worked on algorithms for integrated learning and reasoning (e.g., common sense reasoning)
  • Previously developed a theory of communication in the absence of standards
  • Also broadly interested in theoretical computer science