Hamideza Mohebbi's Home Page

I am a passionate data scientist who loves solving real-world challenges by introducing creative efficient data solutions. I applied my high-performance computing knowledge to real-world problems like detection of cancer fusion and structural variations using DNA/RNA genome data. My past research which is the result of working as a research assistant at the Computer Science Department of the University of Massachusetts Boston. My research interest include, but not limited to:

• Parallel Computation with focus on Nvidia GPUs using CUDA language

• Distance Metric Learning and Object Detection using Deep Learning

• Computational Bioinformatics

This is my CV in PDF.

Education

2014 - 2020: Ph.D in Computer Science, Computer Science Department, University of Massachusetts Boston (UMass Boston).

2007 - 2010: M.Sc. in Software Engineering, Computer Engineering Department, Iran University of Science and Technology,Tehran, Iran.

2003 - 2007: B.Sc. in Software Engineering, Computer Department, Engineering Faculty, Ferdowsi University of Mashhad, Mashhad, Iran.

Projects

• • "A Bidirectional Context-Based Deep Learning Framework for Crater Detection from Both Crater and Non-crater Ends ": This project proposes a new crater detection framework named BCB-Crater-Detection that learns bidirectional context-based features from both crater and non-crater ends. This framework utilizes both craters and its surrounding features using deep convolutional classification and segmentation models to identify efficiently sub-kilometer craters in high-resolution panchromatic images. You can use the following link to read about about this project. Link: https://github.com/mohebbihr/bcb-crater-detection

  • "Learning Weighted Distance Metric From Group Level Information and Its Parallel Implementation ": The objective of this project is to develop a new weighted distance metric learning algorithm that learn group level information from both training and test sets. Our experiments shows that the classification accuracy improved by more than 10%. The parallel implementation of this method using both CPU and GPU can reach more than 3.7 times speedup compare to traditional CPU code in our experiments. You can read more about this project using the following link: https://github.com/mohebbihr/WDM-CUDA

  • "Detection of Chromosomal Structural Variation using Jaccard Distance and Parallel Architecture" : In this work, we proposed a method called TDJD that identifies the location of inter-chromosomal breakpoints corresponding to large scale structural variations, in particular translocations and insertions. To reduce the huge dimension of the search space, we split candidate reads that can be potential breakpoints into windows, and represent the windows as a sequence of binary fingerprints. We then search for the location of the breakpoint in the reference genome using Jaccard distance. You can find source code and more info about this project using the following link: https://github.com/mohebbihr/TDJD

  • "Parallel Berlekamp-Massey Algorithm using GPU": We are performing research on the Berlekamp-Massey algorithm. The first goal is to reach high performance in CPU using SSE and AVX instructions. Another goal is to develop a parallel implementation of BMA algorithm using concurrent kernel execution method in GPU. You can find out more about this project using this link: https://github.com/mohebbihr/Parallel-BMA

  • "A Transparent and Portable Distributed Shared Memory in Virtual Machine Monitor": This was my M.Sc. thesis project under supervision of Dr.Sharifi in Iran University of Science and Technology. The aim of this project, is to produce a distributed shared memory system with good performance, transparency, along with portability while operating system should be kept intact. We achieved this goal by using virtualization technology. In implementation of this project I used the advice of Mr. Cam Macdonell from University of Alberta.

You can read about projects under my supervision, here.

Publications

Thesis and Seminars:

1. HamidReza Mohebbi, "A Transparent and Portable Distributed Shared Memory in Virtual Machine Monitor", M.Sc. Thesis, Iran University of Science and Technology, Under Supervision of Dr. M. Sharifi, 2010. [PDF (English Abstract)] [PDF (Farsi)]

2. HamidReza Mohebbi, "Parallel Programming with MPI", B.Sc. Thesis, Ferdowsi University of Mashhad, Under Supervision of Dr. H. Deldari, 2007.

Journal Papers:

1. ·Hamidreza Mohebbi, Nurit Haspel, Dan Simovici, and Joyce Quach, "Fusion Transcript Detection from RNA-Seq using Jaccard Distance."; 11^th ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (ACM BCB); 2020 (Accepted).

2. · Hamidreza Mohebbi, Wei Ding; "A Bidirectional Context-Based Deep Learning Framework for Crater Detection from Both Crater and Non-crater Ends"; ACM Transactions on Knowledge Discovery from Data (TKDD); 2020; (under review).

3. HamidReza Mohebbi; "Parallel SIMD and GPU implementations of Berlekamp-Massy Algorithm"; International Journal of Parallel Computing; 2018. [Link]

4. HamidReza Mohebbi, Amir Vajdi, Nurit Haspel, Dan Simovici; "Detecting Chromosomal Structural Variation using Jaccard Distance and Parallel Architecture"; IEEE International Conference on Bioinformatics and Biomedicine (BIBM); 2017. [Link]

5. HamidReza Mohebbi, Yang Mu, Wei Ding; "Learning Weighted Distance Metric From Group Level Information and Its Parallel Implementation"; Applied Intelligence Journal, Volume 45, Number 2, 2016. [Link]

6. Hamid Reza Mohebbi, Omid Kashefi, Mohsen Sharifi; “ZIVM: A Zero-Copy Inter-VM Communication Mechanism for Cloud Computing”; Journal of Computer and Information Science; Vol. 4, No.6; November 2011. [PDF]