My name is Hamed Saleh, and I am a PhD student in the CS Department
at the University of Maryland (2018-present). I am working
under the supervision of Prof. Hajiaghayi.
Currently, my research is focused on designing efficient algorithms for solving combinatorial problems -- such as graph problems, classic DPs, and problems on strings -- on sublinear computational models. Sublinear models include Massively Parallel Computation (MPC), Streaming, Dynamic Algorithms, and Sublinear-space Algorithms.
My dissertation is titled "Sublinear Algorithms for Processing Massive Datasets".
Available Resources:
Prior to this, I was employed as a software engineer at the Cafe Bazaar Inc. for two years (2016-2018), and I was a software engineer intern at PageBites Inc. (imo.im) in Summer 2020.
links:
Contact:
hamed [at] cs.umd.edu or
hameelas [at] gmail.com
Misc.
Visiting Researcher at Toyota Technological Institute at Chicago, hosted by Saeed Seddighin, Summer 2021.
Visiting Researcher at Simons Institute for the Theory of Computing, UC Berkeley, Fall 2018.
External Reviewer for conferences COCOON'18, ISAAC'18, WWW'20, FSTTCS’20, SODA'21, ESA'21, and DISC'21.
Software Engineer Intern at PageBites Inc. (imo.im), Summer 2020.
Software Engineer at Cafe Bazaar Inc., 2016-2018.
Dynamic Subset Sum with Truly Sublinear Processing Time [submitted]
me,
Saeed Seddighin
Resources:
∅
Improved Adaptive Massively Parallel Algorithms for Cut Problems [submitted]
MohammadTaghi Hajiaghayi,
Marina Knittel,
Jan Olkowski,
me
Resources:
∅
Massively Parallel Suffix Trees and Applications [submitted]
Jacob Gilbert,
MohammadTaghi Hajiaghayi,
me,
Saeed Seddighin
Resources:
∅
Massively Parallel Constant-round Tree Contraction [ITCS'22]
MohammadTaghi Hajiaghayi,
Marina Knittel,
me,
Hsin-Hao Su
Resources:
Massively Parallel Algorithms for String Matching with Wildcards [SPAA'21]
MohammadTaghi Hajiaghayi,
me,
Saeed Seddighin,
Xiaorui Sun.
Resources:
Computational Analyses of the Electoral College: Campaigning is Hard
But Approximately Manageable [AAAI'21]
Sina Dehghani,
me,
Saeed Seddighin,
Shang-Hua Teng.
Resources:
Streaming and Massively Parallel Algorithms for Edge Coloring [ESA'19]
(also appeared in [DISC'19] as a brief announcement)
Soheil Behnezhad,
Mahsa Derakhshan,
MohammadTaghi Hajiaghayi,
Marina Knittel,
me.
Resources:
Externalities and Fairness [WWW'19]
(also appeared in Social Choice and Welfare journal, Aug 2020)
Mohammad Ghodsi,
me,
Masoud Seddighin.
Resources:
Task Preparation System, tps.
with Kian Mirjalali
tps is a command-line interface for developing IOI/ICPC-like problems,
with a workflow mainly inspired by Polygon.
It has been used to prepare the tasks in IOIs (2017,
2019,
2020,
2021,
2022),
Iranian National Olympiad in Informatics,
and ICPC Asia Region - Tehran site.
There is also a web interface
counter-part implemented for tps.
Kian, my main co-author of this project who kept maintaining the framework since 2017 also published a
technical report
along side other people involved in the TPS ecosystem in Olympiad in Informatics 2019 journal.
Optimizing the Recommendations in Cafe Bazaar's Search using PageRank.
with Sahel Sarcheshmehpour
We incorporated the user activity data in the Search section into the recommendation system for
dynamic app suggestions.
We used PageRank random walks to measure the popularity and the relevance
of an app recommendation.
Resources:
PhyloTrees (github).
with Marina Knittel and Yusuf Alnawakhtha
An implementation of the Neighbor Joining algorithm for constructing
Phylogenetic Trees.
(A project in "Computational Genmoics" course with Prof. Mihai Pop)
HelliJudge, or jux.
with Mahrud Sayrafi
A programming competition judge with focus on security. The judge runs programs in
a minimal chroot "jail" environment based on Fedora to simulate secure sand-boxing in pre-docker era.
IOI Site Administration System, site-admin.
with Amin Khashkhashi Moghaddam
A web-interface to monitor and manage the nodes in a programming contest hall.
site-admin integrates the data of participants and physical nodes,
and it allows you to perform personalized bash or python scripts on the nodes remotely,
and to monitor the progress.
