Hi, I am Mohammad 👋🏻

Biography📖

My name is Mohammad Mashreghi. I’m a four-year EE B.S Candidate at University of Tehran (ECE department), supervised by Prof. Hamed Kebriaei. My interests are mainly in Machine learning and Data science in Finance, Reinforcement Learning,Disturbuted Optimization , and I also love programming. :)

Education🎓

• University of University of Tehran, 2020-Present
  • Degree: B.S
  • Major: Electrical Engineering (Fields: Control theory & Dynamical Systems)
  • GPA: 3.91/4.0
• University of University of Tehran, 2020-Present
  • Major: Minor in Business Management
• Derkhshande Sarraf High School
  • Degree: Diploma
  • Major: Mathematics and Physics
  • GPA: 19.07/20

PUBLICATIONS📚

• Risk Sensitivity in Markov Games and Multi-Agent Reinforcement Learning: A Systematic Review (Presentation beta version)

• Resilient Federated Vision Transformer for Alzheimer’s Disease Prediction with Brain Imaging Data submitted(Presentation)

Research Experience📚

• June 2023 - September 2023: Machine Learning Intern
  • Smart Network Lab
  • Research on designing new voice packages according to the behaviour of Users packages consumption in recent years.
• Feb 2023 - March 2023: Machine Learning Intern
  • Smart Network Lab
  • Reviewed various distributed RL methods applied to traffic signal control, focusing on assumptions like loss functions and system robustness against attacks, and evaluated the effectiveness of each approach in optimizing traffic flow and resilience under adversarial conditions.

Skills💻

• Soft Skills: Team work, Social Communication, Adaptability, Critical Thinking
• Programming Languages: Python, C/C++, MATLAB, Verilog, LaTeX
• Frameworks & Libraries: PyTorch, TensorFlow, PyTorch, Torchattack,Torchvision, scikit-learn, Simulink
• Hardware & System Design: STM32Cube, ModelSim, Intel Quartus Prime, Ppice
• Language: Persian: Native, English: Advanced

Awards and Honors 🎖️

• Top 20% of Class , University of Tehran: Ranked 17th among 128 contestants in Electrical Engineering B.Sc. students, University of Tehran

• Ranked the 8th in control engineering, University of Tehran.

• Full Scholarship from the University of Tehran.

• Iran’s University Entrance Exam , Sanjesh :Ranked 1016th (Top 0.67%) in the National Universities Entrance Examination among more than 155,000 contestants.

Research Interests🔍

• Adversarial ML
• Reinforcement Learning
• Federated RL
• Game Theory
• Optimization
• Analyze Finance Market

CERTIFICATES 🏆

• Using Python to Access Web Data , Coursera

• Advanced Learning Algorithms , Coursera

• Object‑Oriented Data Structures in C++ , Coursera

• Introduction to Git and GitHub , Coursera

• Deep Learning , Neuromatch Academy

PROJECTS EXPERIENCE🚀

• Distributed Cooperative Competitive Multi Agent Reinforcement Learning in Markov Games: Implementiong Q-learning, Actor-critic, Minimax, Belief based algo, Independent Q-learning, and Disturbuted Q-learning, DDPG, MADDPG.

• Robust-Federated-Primal-Dual-Learning-for-Android-Malware-Classification-via-Adversarial-Robustness: In this project, the goal is to achieve robust federated learning for Android malware classification through adversarial robustness(FGSM-PGD).

• Detect fake picture with ML: In this project, we use fake and real pictures of mountains, sea, and forests to detect fake ones.

• Detect fake picture with ML: In this project, we use fake and real pictures of mountains, sea, and forests to detect fake ones.

• Food Hunting: A simple code for getting food faster than others, from students who don’t want their food in university.

• Deep Learning and Neural Network project: MLP, Transfer Learning, Obejct Detection , Image Captioning , Intent Classification , Extractive QA System , Vision Transformer Image Classification

• Sequential Data Processing with RNN:

• Trade bot: A simple bot to buy and sell in forex and get data from tradingview and analyze candles with MACD 12, SMA, EMA 7 with complicated conditions

• Electromagnetic Levitation System Modeling: A simple simulation in Matlab

• Designing an online market with C++ with different facilities :

• Booth Multiplier: A 5 bit booth multiplier implemented with verilog.

• MIPS: An implementation of a MIPS CPU written in Verilog.

• PacMan game: A simple game in CMD

• Buffon’s needle: Buffon’s needle is one of the oldest problems in geometric probability.
• Euler’s number: Uniform sums and Euler’s number
• Banach’s matchbox: Banach’s match problem is a classic problem in probability attributed to Stefan Banach.
• Birthday problem: In probability theory, the birthday problem asks for the probability that, in a set of n randomly chosen people, at least two will share a birthday.
• Image processing: Detecting specific ICs on a PCB from its image
• RTL Circuits: Design a circuit to calculate hyperbolic cosine approximately using its Taylor series
• UT-messenger: a simple messenger made with C++ that works in command line

INTERESTS🎨📚

• 🎧 Listening to podcasts, 📈 Stock trading, ✈️ Travelling, 🌐 Surfing the internet, 🎮 Video Games, 🏓 Ping pong