Explainable AI – one focus area for our new Postdoc Dr. Kyi Thar
In January 2022, Dr. Kyi Thar arrived in Sweden from South Korea, to start his new position as a Postdoc at STC Research Center. His research will focus on the explainable AI, Internet-of-things (IoT), industrial IoT (IIoT), network intelligence, information security, distributed systems.
What is your background?
For my academic journey, after graduating with a Bachelor of Computer Technology degree from the University of Computer Studies Yangon, Myanmar in 2007, I joined as an IT executive and part-time lecturer in Myanmar Egress. Myanmar Egress is a non-profit organization and was founded to promote Myanmar's educational sector, especially for the young generation. As a young enthusiastic learner, I was always looking for opportunities for further studies. Eventually, I was selected for a fully funded president scholarship to the MS leading to Ph.D. program in Computer Engineering at the Networking Intelligence Lab, Kyung Hee University and I graduated in February 2020.
After Ph.D. graduation, I worked as a Postdoctoral Researcher in the Intelligent Networking Lab, Department of Computer Science and Engineering at Kyung Hee University, South Korea. I have collaborated with other organizations in various projects, including the “Institute for Information and Communications Technology Promotion (IITP)” (Korea) and “Electronics and Telecommunications Research Institute (ETRI)” (Korea). I was also collaborating with Trinity Healthcare Enterprise to build an integrated one-stop healthcare solution to improve Myanmar’s healthcare sector. Moreover, I am working with PIT Technology to develop the Agricultural Platform to improve Myanmar’s agricultural sector. The objective is to establish a traceable unified agricultural platform to address information asymmetry, linkage in the value chain system, inconsistency in the quality of produce, impacts of climate change and access to the market of increased income for farmers and stronger sectoral growth in Myanmar.
What is your area of interest?
My research area was content-centric networking and I developed various named-based routing and intelligent caching solutions. I also applied deep learning in intelligent wireless network virtualization and management. The title of my Ph.D. dissertation was “An autonomous deep learning model generation framework for edge computing”, which aims to develop an autonomous system capable of generating the most appropriate domain-specific deep learning model for the edge computing environment. I am also interested in Internet-of-things (IoT), industrial IoT (IIoT), network intelligence, information security, and explainable distributed/federated machine learning research areas.
What is it that you are going to investigate as a postdoc?
As a Postdoctoral Researcher in the Department of Information Systems and Technology at Mid Sweden University, I am going to conduct individual and collaborative research on the explainable AI, Internet-of-things (IoT), industrial IoT (IIoT), network intelligence, information security, distributed systems. Currently, I am also conducting collaborative research on a generalized eXplainable AI-based intrusion detection solution to enhance the detection rate of various types of cyber-attacks.
As part of my research, I am developing a system called Explainable Neural Network Architecture Search (ENAS) to provide explainable edge intelligence solutions for the IIoT system, such as IIoT network management and security. ENAS is an extension of my earlier open-source project done at Kyung Hee University. The ultimate goal of developing ENAS is to possess a global knowledge base constructed in a distributed manner that can be utilized to generate multiple optimized learning models for various purposes such as intrusion detection models for the intelligence intrusion detection systems (IIDS). To achieve the ultimate goal, we aim to utilize the best features of genetic algorithm, meta-learning, asynchronous federated learning, etc.
Why did you choose Mid Sweden University?
I choose Mid Sweden University based on the following facts:
- Sweden is one of the few countries in the world in which it is possible to achieve work-life balance and spend quality time with family and friends or pursuing hobbies.
- The campus is in Sundsvall, which is one of the most beautiful cities in Sweden, and its surroundings are stunning and close to nature.
- Mid Sweden University has a strong collaborative research team, and I want to be part of it.