Department of Computer Engineering • Sharif University of Technology
GPA: / 4.00( / 20.00)
Sep 2026 – Expected 2028
Open to academic collaborations & discussions
Arman Heidari
M.Sc. Student in Artificial Intelligence
Sharif University of Technology • Tehran, Iran
Driven by a deep curiosity for machine learning, neural architectures, and intelligent systems—with reinforcement learning standing as the ultimate frontier to explore.
AI / ML
01 - About Me
Intellectual trajectory, engineering philosophy, and personal curiosities beyond the screen.
The Intellectual Journey
From First-Principles Engineering to Autonomous Agency
My interest in computing started with a simple obsession: understanding how complex systems function from the ground up—how lines of code translate into mathematical transformations and predictable physical behavior. During my undergraduate training in Computer Engineering, I spent years immersed in algorithms, discrete mathematics, and systems programming, developing an enduring respect for first-principles thinking and clean implementation.
Over time, that curiosity naturally drew me toward Artificial Intelligence. What fascinates me most about Machine Learning and Deep Learning is the shift from writing handcrafted, explicit rules to designing architectures that can discover representations directly from data. I find immense satisfaction in bridging theoretical intuition—linear algebra, probability, and optimization dynamics—with concrete, functional code.
To me, Reinforcement Learning represents the ultimate frontier—the "final boss" of AI. The idea of an agent that doesn't merely classify static inputs, but actively explores an environment, learns through trial and error, and formulates strategies over time is deeply compelling. While I am still early in this pursuit and building up the foundational machinery, mastering decision-making under uncertainty and autonomous learning is the summit I'm working toward in my graduate studies at Sharif.
Whether working on machine learning pipelines or studying theory, I prefer substance over hype: building things from first principles, understanding the math before running the code, and never losing the curiosity that started it all.
Core Intellectual Interests
01
Machine Learning & Deep Architectures
Neural representations, generative modeling, representation learning, and modern sequence architectures.
02
Autonomous Systems & RL (The Summit)
Sequential decision-making under uncertainty, policy optimization, and goal-directed agent behaviors.
03
Mathematical Foundations & Optimization
Linear algebra, probability theory, multivariable calculus, and optimization dynamics that make learning possible.
04
Systems Rigor & First Principles
Algorithmic problem-solving, clean code craftsmanship, reproducible experiments, and systems engineering.
