Research

Title: Heating and Cooling Timescales of Transient Brightenings in the Solar Transition Region Loops (2022-Present)

Highlight: Analyzed solar transition region loops using large-scale IRIS slit-jaw images and spectra, identifying and characterizing 71 brightening events. Applied image processing and big data analytics to assess the heating and cooling timescales, statistically comparing these with HYDRAD modeling and classifying loop heating based on the timescale data.

Supervisor: Dr. Shah Md. Bahauddin, Research Faculty and Scientist, Laboratory for Atmospheric and Space Physics (LASP), University of Colorado Boulder. (Personal Web-page)

Title: Context aware gamification based Cognitive Social Computing for user preference selection  (2021-2023)

Highlight: Developed a data-driven gamification framework for personality prediction and personalized recommendations, leveraging large-scale datasets from surveys, gameplay, and social media, and applying BERT and NLP features for advanced feature extraction and analysis.

Supervisor: Dr. Md Saddam Hossain Mukta, Post-doctoral researcher, Software Engineering, LUT School of Engineering Sciences, Lappeenranta- Finland. (Web-page)

Title: Predicting Personalized Preferences in the Metaverse  (2024-2025)

Highlight:

We developed a novel transitive inference model (MV → preference) by integrating the previous two models. This model directly predicts users’ personalized preferences based on their patterns of actions performed on the MV platform, eliminating the need for tedious, manual, questionnaire-based

surveys
Supervisor: Dr. Md Saddam Hossain Mukta, Post-doctoral researcher, Software Engineering, LUT School of Engineering Sciences, Lappeenranta- Finland. (Web-page)

Title: LPWAN Technologies for IIOT and HCI (2023-2024)

Highlight: We conducted a data-driven analysis of various IoT physical layer communication protocols, including NB-IoT, LTE-M, Sigfox, LoRa, Z-Wave, and RedCap. Our evaluation focused on key metrics such as power consumption, transmission range, and application performance to assess the effectiveness of each protocol. Additionally, we analyzed the underlying network architecture to identify the strengths and weaknesses of each technology. This comprehensive approach allowed us to determine the optimal protocol for Industrial Internet of Things (IIoT) and Human-Computer Interaction (HCI) applications. Our findings contribute valuable insights for selecting suitable communication protocols in these domains.

Supervisor: Dr. Arshia Khan, Professor & Director of Graduate Studies, University of Minnesota Duluth. (Personal Web-page)

Title: GPU Driven Deep Learning based Model Analysis (2023-2024)

Highlight: We implemented a data-driven comparison of GPU and CPU performance in deep learning, aiming to optimize processing efficiency. Our analysis focused on key metrics, including runtime, inference time, and memory usage, using large datasets to ensure comprehensive evaluation. By conducting this comparison, we identified the strengths and weaknesses of each processing unit, providing valuable insights for selecting the appropriate hardware for deep learning tasks. This project contributed to improving the overall performance and efficiency of deep learning models in various applications.

Supervisor: Dr. Abdulaziz Tabbakh, Asst. Professor, King Fahd University of Petroleum and Minerals (KFUPM). (Web-page)

Title: Sustainable AI Framework for Green AI (2023-2024)

Highlight: This research positions Green AI as a crucial direction for future research and development. It proposes a comprehensive framework for understanding, implementing, and advancing sustainable AI practices. We provide an overview of Green AI, highlighting its significance and current state regarding AI’s energy consumption and environmental impact. The paper explores sustainable AI techniques, such as model optimization methods, and the

 development of efficient algorithms  

Supervisor: Dr. Abdulaziz Tabbakh, Asst. Professor, King Fahd University of Petroleum and Minerals (KFUPM). (Web-page)

Title: A Data Driven Approach to Analysis of the Effect of Pandemic on Society (2024-Present)

Highlight: We performed a data-driven analysis of AI applications during the COVID-19 pandemic through a comprehensive literature survey, examining over 200 peer-reviewed studies. Our focus was on knowledge discovery and AI-based decision-making processes that contributed to societal impact, particularly in healthcare and public safety. We quantified the effectiveness of various AI solutions by analyzing their outcomes and real-world applications. This analysis provided valuable insights into how AI technologies were leveraged to address the challenges posed by the pandemic.

Supervisor: Dr. A.K.M. Muzahidul Islam, Professor, United International University, Bangladesh. (Web-page)
Co-Supervisor: Dr. Thanh Thi Nguyen, Associate Professor
Department of Data Science and AI, Monash University, Australia (Personal webpage)

Smart Railway Crossing Monitoring System (2013)
Institute: Ahsanullah University of Science and Technology, Dhaka              

Highlight : We developed and implemented an automated railway crossing monitoring and control system utilizing embedded systems integrated with IoT protocols. This system enabled real-time data collection and remote management of railway crossings, significantly enhancing operational efficiency. By automating critical processes, the prototype minimized human error and improved safety at railway intersections. The IoT integration allowed for seamless communication between devices, providing timely updates for better decision-making and control. This project demonstrated the potential of IoT in improving the reliability and safety of railway crossing management systems. 

Project name : B2C Automation and billing Solution for Service Provider - Hydra  (2020-2021)
Institute : Link3 Technologies Limited, Dhaka

Highlight: The user effectively tested and integrated an advanced end-to-end OSS system for 150,000 internet clients, ensuring seamless interoperability between broadband networks, EPC, soft switches, OTT services, IPTV middleware, and IoT platforms. This integration streamlined communication between diverse platforms, resulting in enhanced operational efficiency and improved service quality. By optimizing the system, the user facilitated smooth delivery across multiple technologies, improving scalability and reliability. Their efforts contributed to more efficient network management and better overall user experience.

Key Features Integration :

•Implemented triple play and quad play bundles: data, TV/OTT, telephony, payment
•Billing System integration for 0.15 million subscribers.
•Integrated with quotas / FUP
•Integrated with time intervals