About   Education   Research   Publication   Project   Work   Award   Skill  
AS MD HABUBULLAH

I am doing a Doctor of Philosophy (PhD) at Xidian University, focusing on Deep Reinforcement Learning (DRL), Supervised by Wang Shuzhen. Demonstrated leadership at Autofficina Internazionale s.a.s ( InternazionaleAuto ), skilled in Vehicle Automation system development and project management.

With five years of professional job experience in Senior Software Engineering and a strong academic background (ML, DL, and DRL), I'm looking for a part time research opportunity to contribute to the FinTech field. My current research/project is Financial Deep Reinforcement Learning (FinDRL), funded by the Shaanxi State Key Lab "Institute of Big Data and Visual Intelligence".

Education

Sep 2025—PhD Computer Science and Technology
July 2029Xidian University
Research Topic: Deep Reinforcement Learning in FinTech (FinDRL Approach)

Sep 2022—MEng Computer Science and Technology
July 2025Xidian University
Overall Grade: 91.84%
Thesis: Stock Market Price Forecasting Using Deep Reinforcement Learning Techniques (FinDRL Approach)

Sep 2018—BE. Computer Science and Technology
July 2022Yunnan University
Overall Grade: 80%
Thesis: Cross Border E-Commerce Solution (B2B & B2C)

Ongoing Research Highlight
End-to-End Stock Market Technical Indicator Parameters Optimization and Performance Analysis: A Dynamic Approach
- Investigate dynamic optimization methods to enhance indicator-based trading strategies.
This research framework revolutionizes technical analysis by automating parameter optimization and rigorous validation. It establishes a new standard for evaluating trading strategies, ensuring all parameters and indicators are thoroughly tested, enabling smarter and more profitable decisions.
Paper Code Simple explanatin
Deep Neural Network Based Stock Indicator Optimization and Predictive Analysis (SIOPA) Framework for Financial Time Series
- Develops a novel architecture for hyperparameter tuning in financial time series data, yielding improved predictive accuracy for market trends.
Predicting Optimal Indicator Parameters with DNNs Develop a deep neural network to classify stock market indicator parameters as worst, bad, good, better, best, or special. Using historical data and backtesting metrics (e.g., Sharpe ratio, max drawdown), this model streamlines parameter optimization for improved trading performance.
Paper Code Simple explanatin
FinDRL: Financial Deep Reinforcement Learning: A Perfect Way to Manage Stock Trading Portfolios.
- Proposes an advanced reinforcement learning framework for adaptive portfolio rebalancing.
FinDRL applies reinforcement learning to optimize financial strategies, enabling adaptive decision-making in dynamic markets. It leverages advanced algorithms to maximize returns while managing risk, revolutionizing investment and trading approaches.
Paper Code Simple explanatin
FinDRL – An Integrated Framework for Stock Breakout Trading Strategy and Portfolio Risk Management: A Machine Learning Hybrid Model Approach
- Breakout Trading Strategy, Machine Learning in Finance, KMeans Clustering, LightGBM Classifier, Portfolio Risk Management
This research is a hybrid machine learning framework for breakout trading, combining unsupervised pattern discovery and supervised prediction with dynamic risk management to outperform traditional rule-based strategies.
Paper Code Simple explanatin
Publication

Research on the Impact of Fiscal and Tax Policies And Social Responsibility on the Financial Performance of Listed Chinese Agricultural Companies.
Md Mostafijur Rahman, Mohsina Khatun and, Abu Sayem Md Habibullah
International Journal of Law, Humanities & Social Science
Paper

An empirical study on corporate social responsibility, environmental regulation and financial performance-Based on heavy pollution industry
Md Mostafijur Rahman, Yanchun Zhu and Abu Sayem Md Habibullah
INTERNATIONAL JOURNAL OF BUSINESS, SOCIAL AND SCIENTIFIC RESEARCH
Paper

Top Project

Feb 2022Cross Border E-Commerce Solution (B2B&B2C).
• Associated with Yunnan University.
• Stack: - Docker, Scaffold, Kubernetes, DigitalOcean, JavaScript, TypeScript, Microservice Architecture

May 2021Customer Relation Management (CRM) & Task Automation System.
• Associated with INTERNAZIONALEAUTO.
• Stack: Git, Docker, Kubernetes, Heroku, AWS, MongoDB, Redis, PostgreSQL, NextJs, and TensorFlow.

Jan 2021Smart Charging Station with ESP32, and RaspberryPI.
• Associated with Yunnan University.
• Stack: AWS, ESP32, Arduino, C, C++, NextJS, JavaScript, TypeScript, PostgreSQL, Raspberry Pi, Arduino (ESP32), Microservice Architecture

Selected Award

May 2022National Undergraduate E-commerce "Innovation, Creativity and Entrepreneurship" Challenge - Position: 12th
• Awarded for: China Bangladesh E-commerce integrating transportation and marketing.
• Issued by Yunnan University.

Jan 2022China-Bangladesh e-commerce mobile platform integrated with transportation and sales - Position: 12th
• Awarded for: The work "Sino-Bangladesh Express", a China-Bangladesh e-commerce mobile platform integrated with transportation and sales, challenged Zhaizhong Ronghua in the 12th National College Student E-commerce "Innovation, Creativity and Entrepreneurship" at Yunnan University
• Issued by Sino-Bangladesh Express.

Work Experience

May 2021Autofficina Internazionale s.a.s
PresentSoftware Engineer - Remote | May 2021 - Present
▪ SaaS application offering services: Customer appointment management, Marketing system, Dynamic page builder, URL shortener, and Telematics implementation to enhance business operations.
▪ Documented improvements, highlighting significant reduction in data size and storage. Reported a 33.33% improvement in storage efficiency.
▪ SaaS application offering services: Customer appointment management, Marketing system, Dynamic page builder, URL shortener, and Telematics implementation to enhance business operations.
▪ Successfully increased customer satisfaction by 10% by building a Deep Neural Networks model to provide solutions for 14 vehicle problems via our company's services.
▪ Improved server query response by 15% through engineering best practices with Microservices architecture, DCS, and CI/CD pipeline.

Jan 2021BreezeBangladesh LTD
Apr 2023IT Team Instructor - Remote | Jan 2021 - Apr 2023
▪ I offered technical assistance to the IT team beyond established working hours to utilize my expertise to troubleshoot and resolve customer issues.
▪ Develop customer trust by evaluating the criteria and optimizing customer guidance.
▪ Software development pipelines through CI/CD, increasing the development team's speed by 20%.

Skill

Language/Frameworks/Tools/Concepts
▪ C++, Python, and TS
▪ Docker, Kubernetes, Kubeflow, AWS, HTTP, and gRPC Protocol
▪ NodeJs, ReactJs, TensorFlow, PyTorch, MLflow, and Deep Learning