Interview with Sharia Arfin Tanim
In this episode of "Young Scientists Interviews", we had the privilege of speaking with Sharia Arfin Tanim, a researcher at the Learnify Research Lab (LRL) and a Research Assistant (RA) at the Advanced Machine Intelligence Research Lab (AMIR Lab). Below is the full transcript of our conversation:
Tell us a bit about yourself.
I completed my BSc in Computer Science from American International University-Bangladesh (AIUB) in August 2024 and am the founder of the Learnify Research Lab (LRL).
What are your research interests?
My primary research focus is on decentralized learning methods, particularly in the context of federated learning. This approach enables artificial intelligence (AI) models to train on multiple devices, such as smartphones or computers, without centralizing data. By keeping data on local devices and sharing only model updates, federated learning enhances privacy and improves data security.
Additionally, my research spans areas like large language models, computer vision, and pattern recognition, aiming to develop AI systems that understand human language, analyze visual data, and identify patterns in complex datasets. Specifically, I have applied AI to analyze medical and agricultural images, helping improve diagnostics and crop management. This interdisciplinary research contributes to building more efficient, privacy-preserving AI systems with significant applications in healthcare and agriculture.
How does your research benefit society?
Through my research, I aim to address critical societal challenges:
- Privacy Preservation: Federated learning ensures that personal data remains on devices, reducing the risk of data breaches and misuse.
- Improved Healthcare: AI can analyze medical images, assisting doctors in diagnosing diseases more accurately and efficiently, leading to better treatment outcomes.
- Agricultural Advancement: AI helps farmers by analyzing crop images, identifying diseases, managing resources, and predicting yields more effectively.
- Enhanced AI Tools: Advances in language models and computer vision make AI more capable of interacting with people and solving everyday problems, leading to innovations in virtual assistants and smart educational tools.
Overall, my research contributes to creating more secure, efficient, and impactful AI systems that benefit society across multiple domains.
Can you share any specific research experiences?
One of my early research projects focused on using pre-trained deep learning models for disease classification. This involved collecting medical data and addressing privacy challenges. Since medical images often contain sensitive patient information, ensuring privacy and securing the data-sharing process were critical tasks.
Another challenge was the availability of data. For instance, if a disease is more prevalent in Africa but less common in Asia, there can be a huge imbalance in the data required for effective model training. This led me to explore decentralized approaches, where federated learning emerged as a promising solution. Federated learning enables decentralized model training, eliminating the need for centralizing data and reducing privacy concerns.
Currently, I am addressing a known issue in federated learningānon-IID (non-independent and identically distributed) dataāto unlock its full potential for medical applications while maintaining data privacy.
What qualities do you think are essential for a scientist?
As a researcher, I believe that patience and dedication are the most critical qualities to cultivate. Once these traits are developed, innovative thinking, sharp observation, and the ability to solve complex problems naturally follow.
What message do you have for aspiring scientists in Bangladesh?
Always stay curious and never stop asking questions. Patience and persistence are crucial, as challenges are an inevitable part of research. Work ethically and maintain collaboration with others. Nurture your passion while also taking care of yourself.
As Marie Curie once said, "Nothing in life is to be feared; it is only to be understood." Your work is valuableākeep moving forward! Best of luck.
Contact Information
- LinkedIn: www.linkedin.com/in/learnifyresearchlab
- Website: learnifyresearchlab.com
- ResearchGate: ResearchGate/learnifyresearchlab
- Email: shariaarfin096@gmail.com | official@learnifyresearchlab.com