Kooshan Maleki

Hello! I'm Kooshan Maleki, a senior undergraduate student in Computer Engineering at Amirkabir University of Technology . I'm passionate about quantum computing and machine learning, with research focusing on hybrid quantum-classical algorithms. My work includes enhancing the HHL algorithm through deep learning-based Pauli decomposition and optimizing Quantum Neural Networks using circuit cutting and evolutionary algorithms. I'm interested in exploring how quantum computing can revolutionize machine learning and solve complex optimization problems.

Location: Tehran, Iran

Education

B.Sc. in Computer Engineering

Institution: Amirkabir University of Technology (2021 – Present)

GPA: 18.84/20 (4/4), Top 5%

Bachelor Thesis: Analysis and Comparison of the HHL Algorithm with Classical Methods for Solving Linear Systems and Its Application in Linear Regression

Diploma

Institution: Allameh Helli High School (NODET) (2017 – 2021)

GPA: 19.40/20

Research Experience

Advisor: Dr. Alberto Marchisio (eBRAIN Lab at NYUAD)

Description: Research on developing a hybrid Quantum Neural Network (QNN) optimization framework that combines circuit cutting techniques with evolutionary algorithms to enhance quantum machine learning model performance and efficiency.

Exploring Neural Architecture Search (NAS) methodologies to automatically discover optimal quantum neural network architectures.
Investigating circuit cutting techniques to reduce quantum circuit complexity and improve scalability.
Implementing evolutionary algorithms for efficient architecture search and optimization.
Analyzing trade-offs between model accuracy, computational cost, and quantum resource constraints.
Developing hybrid quantum-classical frameworks for automated architecture search and evaluation.

Advisor: Prof. Negar Ashari (Amirkabir University of Technology)

Investigating methods to enhance the Hybrid HHL (Harrow-Hassidim-Lloyd) algorithm through deep learning approaches.
Developing deep learning-based Pauli decomposition techniques to improve quantum algorithm efficiency.
Analyzing the integration of classical machine learning with quantum algorithms for enhanced performance.
Exploring applications of enhanced HHL algorithm in solving linear systems and quantum machine learning tasks.

Industrial Experience

Website: https://mfp.co.ir

Developed a machine vision system for analyzing heart ultrasound images to enhance the accuracy and efficiency of medical diagnostics.
Implemented AI algorithms using OpenCV and machine learning techniques to automate image segmentation and contour detection in medical imaging.
Collaborated with Shahid Rajaei Hospital to collect and process real-world medical data, ensuring high accuracy in AI-driven analysis.
Worked with technologies like PySide6 and Matplotlib to build interactive user interfaces and visualize medical data.
Contributed to localizing cardiac ultrasound equipment, reducing dependency on foreign technologies through innovative AI solutions.
Developed algorithms for precise measurement of heart muscle strain and ejection fraction (EF) using advanced image processing techniques.

Selected Projects

Quantum Machine Learning (2024)

Amirkabir University of Technology

Developed hybrid quantum-classical models using Qiskit, leveraging variational algorithms to optimize model parameters for enhanced performance. Implemented amplitude encoding techniques based on Motonen's methodologies to efficiently represent classical data within quantum states. Utilized Quantum Fourier Transform (QFT) and Inverse Quantum Fourier Transform (IQFT) to improve feature extraction and facilitate complex data analysis. Conducted experiments to compare the efficacy of quantum-enhanced models against traditional classical approaches, demonstrating potential advantages in specific machine learning tasks.

Repo Loading...
Kolmogorov-Arnold Neural Networks (2024)

Personal Project

Implemented a neural network from scratch, exploring spline interpolation techniques to enhance network convergence and performance. Developed custom activation functions and backpropagation algorithms to optimize learning efficiency. Utilized spline interpolation for smoother gradient updates, reducing oscillations and improving the stability of the training process. Conducted extensive testing on various datasets to validate the robustness and scalability of the implemented neural network architecture. Documented the development process and findings, contributing to a deeper understanding of the interplay between spline interpolation and neural network optimization.

Repo Loading...
A Hybrid QNN Optimization Framework with Circuit Cutting and Evolutionary Algorithms (2025)

Research with Dr. Alberto Marchisio (eBRAIN Lab at NYUAD)

Developing a hybrid Quantum Neural Network (QNN) optimization framework that combines circuit cutting techniques with evolutionary algorithms to enhance quantum machine learning performance. Exploring Neural Architecture Search (NAS) methodologies to automatically discover optimal quantum neural network architectures, reducing circuit complexity and improving scalability through advanced optimization techniques.

Teaching Experience

Advanced Programming (under supervision of Dr. Taromi Rad) Fall 2025
Software Engineering II (under supervision of Dr. Gohari) Fall 2025
Special Topics in Quantum Computing (Graduate Course, Head TA, under supervision of Dr. Negar Ashari Astani) Spring 2025
Microprocessors and Assembly (Head TA, under supervision of Dr. Farbeh) Spring 2025
Computer Networks (under supervision of Dr. Sabaei) Fall 2024, Spring 2025
Applied Linear Algebra (under supervision of Dr. Nazerfard) Fall 2024
Computer Architecture (under supervision of Dr. Zarandi) Spring 2024, Spring 2025
Algorithm Design (under supervision of Dr. Dolati Malekabad) Spring 2023
Logic Circuits (under supervision of Dr. Sedighi, Dr. Saheb Zamani) Spring 2023

Trained students for the Iranian Physics Olympiad, focusing on deep conceptual understanding.

Certifications and Specialized Courses

Organizer: NYU Abu Dhabi (International Hackathon for Social Good in the Arab World)

Focus: Quantum Computing, Quantum Machine Learning

NYUAD Hackathon 2025 Certificate — Certificate

NYUAD Hackathon Credential — LinkedIn — Credential

Institution: The Coding School (QxQ)

Gained foundational knowledge of quantum computing concepts such as qubits, superposition, and entanglement.
Implemented basic quantum algorithms like the Deutsch-Jozsa algorithm using quantum simulators.

Learned key concepts in supervised learning, including regression, classification, and regularization techniques.
Worked on hands-on projects, using libraries like Scikit-Learn to implement machine learning models.

Supervised ML Certificate — Course Certificate

Mastered the fundamentals of deep learning, focusing on neural network architectures and optimization techniques.
Implemented deep learning models using TensorFlow and Keras for image and text classification tasks.

Neural Networks & Deep Learning Certificate — Course Certificate

Honors and Awards

Accepted for participation in the NYUAD Hackathon 2025 working on Quantum Algorithms utilized for social good.
Achieved Top 1% out of over 142,000 participants in the nationwide Iranian Universities Entrance Examination (2021).
Awarded the QubitXQubit course scholarship held by MIT and Stanford Ph.D.
Attended the 32nd Physics Olympiad of Iran.
Winner of the 2017 SamCode Programming Competition in Data Analysis, excelling among 200 NODET students.

Skills and Competencies

Programming Languages: Python, Java, C, C++, Verilog, VHDL
Libraries/Frameworks: PyTorch, TensorFlow, NumPy, Pandas, PyGame, Cirq, Qiskit, Qiskit Aer, Scikit-Learn
Tools/Platforms: GitHub, PySide6, OpenCV, Matplotlib, Kubernetes, Docker

Languages

Speaking: Fluent
Reading: Fluent
Writing: Fluent

Mother tongue (native proficiency).

Hobbies and Interests

Hiking: Enjoy exploring nature and hiking in different terrains.
Tennis: Competed in amateur tennis tournaments and held a tennis certification.
Cooking: Passionate about culinary arts and experimenting with new recipes. Certified in Cooking.
Video Games: Enthusiast of various video game genres, especially strategy and adventure games.