Seyed Ali Hosseini

Supervisor: Prof. Sven Tomforde Description: - Conducted a six-month visiting research project on adaptive state representation methods for deep reinforcement learning in energy market trading

Responsibilities: • Processed and analyzed high-frequency (tick-by-tick) energy market data in real-time • Developed machine learning and deep learning models for pattern recognition and feature extraction from stochastic market data • Implemented unsupervised learning techniques for anomaly detection in energy trading signals • Designed and improved algorithmic trading strategies using AI-driven approaches • Trained deep reinforcement learning agents for real-time automated trading • Built deep learning mo

Responsibilities: • Performed data analysis and preprocessing to support AI model development • Developed Conditional Generative Adversarial Networks (cGANs) for image generation tasks • Applied reinforcement learning methods to optimize architectural design processes • Built and deployed AI-driven applications using the Django framework • Implemented CI/CD pipelines using GitLab CI/CD for automated deployment on virtual servers Skills Covered: • Generative Models (GANs, Conditional GANs) • Reinforcement Le

Research Assistant at Alzahra University, Tehran During my Military Service, I was honored to serve as a Research Assistant at Alzahra University, Tehran, owing to my high final grade in my Master’s degree. My primary responsibility was to develop a user-friendly software solution for the university's research section. This software aimed to streamline the reporting process for past and existing contracts and to facilitate monthly and yearly financial reporting. Key Achievements: Data Collection & Databas

An A2C(Actor-critic) algorithm was implemented to trade automatically in Forex. In addition, some other ML-based algorithms were written to optimize data preprocessing in the financial market

Subject: Deep Reinforcement Learning and Deep Learning for Quantitative Trading in Energy Markets Description: • Developed advanced models for pattern detection in financial time series using novel data analysis techniques • Designed and implemented a new Generative Adversarial Network architecture

Honor: highest-ranked student in the master's degree Subject Covered: Reinforcement learning-based controllers, Advance Automatic Controllers, Intelligence Control, Deep learning Models, Fault-tolerant Control, Optimization, and System Identification. Thesis: Design of fault-tolerant control system

Subject Covered: Programmable Logic Control(PLC), Modern Control, Linear Control. Non-Linear Control, Micro-Controllers, Fuzzy Controller Final Project: Design a Fuzzy controller for adjusting the Temperature of a Tank Supervisor: Dr. Mehdi Shahbazian Thesis Grade: 19.5/20