BIOGRAPHY
Entrepreneur and PhD researcher focusing on computational algorithms, with applications in quantitative finance. IEEE publication author with experience teaching both online and in-person.
EDUCATION
PhD Researcher
University of Hertfordshire
YEARS OF EXPERIENCE
+8
PUBLICATIONS
+2
CERTIFICATIONS
+5
Skills
My favourite skillsI learned at the university
Linux
Intermediate
Python
Intermediate
Pytorch
Intermediate
Deep
Learning
Intermediate
LaTex
Intermediate
C++
IntermediateI learned at work
Next.js
Intermediate
Node.js
Intermediate
Git
Intermediate
HTML
Intermediate
CSS
Basic
SQL
IntermediateExperiences
My lovely experiencesAcademic
Graduate Research Assistant
University of TehranAs a master's student at University of Tehran, I conducted extensive research in distributed systems, parallel computing, and machine learning under the mentorship of Professor Abdorreza Torabi. My efforts resulted in the publication of two research papers:
- 1- L-DATR: A Limited-Memory Distributed Asynchronous Trust-Region Method
- 2- Design and Implementation of an Efficient Parallel Algorithm for Sparse Principal Component Analysis
For the L-DATR paper, I collaborated closely and effectively with Saeed Soori, who holds a Ph.D. in Computer Science from the University of Toronto. This collaboration further enriched my research experience.
Graduate Teaching Assistant
University of Tehran As a Graduate Teaching Assistant at University of Tehran, I contributed to several advanced courses for master’s students, enhancing my ability to explain complex topics. For the Machine Learning course, supervised by Professor Ali Fahim, I facilitated practical labs and deepened my expertise in guiding students through real-world applications. In the Randomized Algorithms course, under Professor Ali Moeini, I developed engaging sessions that simplified probabilistic concepts, boosting my confidence in teaching theoretical material. Additionally, for the Parallel Computing course with Professor Abdorreza Torabi, I supported hands-on coding labs, strengthening my skills in managing technical instruction.Work
Founder
Gatpier Due to my passion for entrepreneurship, I founded a startup called Gatpier. We offered various services, including consulting for companies on leveraging deep learning in their operations, setting up and maintaining company networks, ensuring network security, and teaching practical subjects such as deep learning, data science, optimization, networking, and programming. As the founder, I faced numerous challenges like team management, coordinating with professors from the University of Tehran to teach at our company, ensuring the quality of our educational courses, negotiating with other companies, and overcoming various other obstacles. These experiences significantly enhanced my management skills and problem-solving abilities.Python Programmer
Hamrahe Aval (MCI) I worked as a programmer at Hamrahe Aval, the first and largest mobile operator in Iran, where I built and refined my Python expertise across multiple projects. My role involved developing and optimizing code to support mobile network operations, adhering to strict network protocols, and collaborating with teams to integrate solutions. I also contributed to enhancing system efficiency by automating processes and tackled challenges in mobile network security, gaining practical insights into large-scale telecommunications infrastructure.Interests
My recent research interests
Deep
Learning
Exploring optimisation techniques to improve the efficiency and performance of Deep Learning models.
Distributed
Systems
Building scalable distributed systems that handle complex tasks efficiently, reliably, and across multiple computing resources.
Quantitative Finance
Applying mathematical modelling and computational methods to financial markets and quantitative analysis.
Publications
Recent publicationsIEEE/ACIS 24th International Conference on Computer and Information Science
L-DATR: A Limited-Memory Distributed Asynchronous Trust-Region Method
Abstract A distributed approach is proposed in this work to solve large-scale optimization problems, called L-DATR, under the master/worker communication model. L-DATR is a distributed limited-memory trust-region method that allows worker nodes to perform asynchronous computations. Our method dynamically adjusts the step size and direction using trust-region strategies to improve stability and convergence. To our knowledge, this is the first implementation of a distributed trust-region limited memory quasi-Newton method with robust handling of asynchronous updates and non-uniform delays between nodes. Our method is communication-efficient because it communicates only vectors of the dimension of the decision variable. Our numerical experiments match our theoretical results and showcase significant stability improvements compared to state-of-the-art distributed algorithms.(PDF, Code)
IEEE/ACIS 24th International Conference on Computer and Information Science
Design and Implementation of an Efficient Parallel Algorithm for Sparse Principal Component Analysis
Abstract Sparse matrix computations are an important class of algorithms. One of the important topics in this field is SPCA (Sparse Principal Component Analysis), a variant of PCA. SPCA is used to compute the principal components of a matrix. There are various methods for computing the sparse principal components of a dataset. One of them is the congradU (Conditional gradient algorithm with unit step size) method, which is an iterative approach. This method performs a matrix-vector multiplication at each iteration of its execution process. Therefore, we need to accelerate the multiplication operation. In this regard, we propose a parallel algorithm for the congradU method that uses a master/worker model to distribute the rows of the matrix among the cores or processors in a manner that ensures an appropriate workload distribution between them. By optimizing the workload distribution among processors, we can reduce the overall execution time of operations. The proposed algorithm has been tested on randomly generated matrices with different sizes and sparsity percentages. We compare the time to find the first principal component using the proposed algorithm and SVD algorithm. It was observed that by increasing the size and sparsity percentage of the matrix, the proposed algorithm finds the first principal component faster than the SVD algorithm. Also, we compare the time of the multiplication operation in one iteration of the proposed algorithm and the dot operator (in Python), and we observe that with increasing the percentage of sparsity, the proposed algorithm performs better than the dot operator.(PDF, Code)
Certifications
Some Certifications
