Master's student & researcher
KU Leuven · eBRAIN Lab, NYU Abu Dhabi
I'm Kooshan Maleki, a Master of Engineering: Computer Science student at KU Leuven in Belgium. I work on hybrid quantum-classical machine learning, the part of quantum computing where the interesting question is not whether a circuit runs, but whether it earns its place next to a classical model.
I'm a researcher at the eBRAIN Lab at NYU Abu Dhabi, with Prof. Muhammad Shafique and Dr. Alberto Marchisio. There I built QNAS, a multi-objective neural architecture search framework for quantum neural networks that trades accuracy against circuit cost and wire-cutting partitionability instead of chasing accuracy alone.
I also work at QuCAL with Prof. Negar Ashari Astani, on making the Hybrid HHL algorithm practical through deep learning-based Pauli decomposition. My bachelor's thesis at Amirkabir University of Technology was the honest version of that question: how the HHL algorithm actually compares to classical linear solvers, and what that means for linear regression.
Teaching is the other half. I've been a teaching assistant at Amirkabir since 2021 across nine courses, among them advanced programming, computer architecture, networks and algorithms, and head TA for the graduate quantum computing course. Before that I coached physics olympiad students at my old high school. The full CV has the rest: industry work on cardiac ultrasound imaging, projects, and awards.
A multi-objective framework for automated architecture search over hybrid quantum-classical neural networks. NSGA-II jointly optimizes accuracy, circuit cost and wire-cutting partitionability, with checkpoint-based correlation analysis to stop unpromising candidates early.
