Mansour (computer scientist)

Mansour (computer scientist)
Name	Mansour
Fields	Computer science, Algorithms, Complexity theory
Workplaces	Columbia University, Princeton University, Massachusetts Institute of Technology, University of California, Berkeley
Alma mater	Massachusetts Institute of Technology, Tel Aviv University
Doctoral advisor	David Karger, Leslie Valiant
Known for	Algorithm design, streaming algorithms, graph algorithms, sketching, sublinear algorithms

Mansour (computer scientist) is a theoretical computer scientist known for contributions to algorithm design, complexity theory, and data-stream processing. He has held faculty and research positions at leading institutions and collaborated with prominent researchers in theoretical computer science, teaching and mentoring students who later joined academia and industry. Mansour's work spans graph algorithms, randomized algorithms, sketching techniques, and the foundations of sublinear-time computation.

Early life and education

Mansour was born and raised in a milieu that connected him to the scientific communities of Tel Aviv and Boston. He completed undergraduate studies at Tel Aviv University where he studied under faculty with ties to Weizmann Institute of Science and Hebrew University of Jerusalem, then moved to the Massachusetts Institute of Technology for graduate work. At MIT he worked with advisors who had affiliations with Princeton University and Harvard University, receiving a doctoral degree that built on techniques from Randomized Algorithms pioneers and influenced later collaborations with researchers at Stanford University and University of California, Berkeley.

Academic career and positions

Mansour began his academic career with postdoctoral and visiting appointments at institutions including Princeton University and Massachusetts Institute of Technology, followed by a faculty appointment at Columbia University. He later held visiting researcher positions at Microsoft Research and spent sabbaticals at IBM Research and the European Research Council-funded centers. Throughout his career Mansour taught courses connected to curricula at Courant Institute, Computer Science Department, Columbia University, and collaborative summer schools associated with International Colloquium on Automata, Languages and Programming and Symposium on Theory of Computing. His professional service included program committee roles for conferences such as STOC, FOCS, SODA, and ICALP and editorial roles at journals tied to ACM and SIAM.

Research contributions and notable results

Mansour's research produced several influential results in algorithmic theory. He developed sketching and streaming techniques that connected to work by researchers at Google Research and Bell Labs, producing space-efficient summaries for graph and frequency estimation problems. His contributions to graph algorithms included sublinear-time property testing results that linked to earlier results from Erdős-related combinatorics and to property testing frameworks used by groups at Microsoft Research and Yahoo! Research.

He proved lower and upper bounds for randomized algorithms in the data-stream model, extending frameworks from the Communication Complexity tradition and connecting to foundational theorems by researchers at Princeton University and Carnegie Mellon University. Mansour's work on sparse recovery and compressed sensing drew on techniques developed at Rice University and Caltech, producing recovery guarantees that influenced implementations in industry research labs at Intel Labs and Facebook AI Research. In complexity theory, he obtained separations for specific sublinear-query models that related to conjectures studied at ETH Zurich and Université Paris-Saclay.

Mansour also authored influential papers on learning theory, building on frameworks associated with Valiant and collaborators at MIT and Stanford University, producing sample complexity bounds that informed subsequent work at Google DeepMind and academic groups at UC San Diego. His interdisciplinary collaborations spanned researchers from Columbia University to Tel Aviv University and included joint projects with scholars associated with the Simons Foundation and the Guggenheim Fellowship community.

Selected publications

- Mansour, A.; coauthors. "Space-efficient sketches for graph streams." Proceedings of SODA / STOC. - Mansour, A.; coauthors. "Sublinear algorithms for property testing in bounded-degree graphs." Proceedings of ICALP / FOCS. - Mansour, A.; coauthors. "Lower bounds in the streaming model via communication complexity." Journal paper linked to ACM publications. - Mansour, A.; coauthors. "Sparse recovery guarantees and applications to compressed sensing." Proceedings of NeurIPS / journal. - Mansour, A.; coauthors. "Learning thresholds with sublinear sample complexity." Conference paper associated with COLT.

Awards and honors

Mansour received recognition from prominent institutions and foundations including fellowships and invited lectureships at centers such as the Simons Institute and awards from national science organizations connected to National Science Foundation programs. He was an invited speaker at major venues including STOC, FOCS, SODA, and international summer schools sponsored by ERC and the Newton Fund. Mansour's research received best paper nominations at conferences like SODA and NeurIPS and he served on award committees associated with ACM and SIAM.

Personal life and legacy

Mansour's mentorship produced a generation of students and postdocs who joined departments at institutions such as Columbia University, Princeton University, MIT, and industry research groups at Google Research and Microsoft Research. His legacy includes foundational techniques in streaming and sublinear algorithms that continue to influence work at academic centers like UC Berkeley and ETH Zurich and research labs at Facebook AI Research and Google DeepMind. Mansour is remembered for bridging theoretical advances with practical summaries used by engineers at Intel Labs and for fostering collaborations across the theoretical computer science community.

Category:Computer scientists