A graduate student of applied data science (Supervisor: Simon Sung-Il Woo), Dae-Young Yoon's M.A. dissertation was published in the best academic society in the field of machine learning and information management, the Conference on Information and Knowledge Management (CIKM) 2020 (BK Computer Science IF=3). (22.1% of adoption rate)
Title: "Who is Delivering My Food? Detecting Food Delivery Abusers using Variational Reward Inference Networks"
In this paper, a method of novelty detection was proposed to solve abusing cases that cause numerous problems such as violations of the paid-in-full transportation law in the food delivery market. The designed methods are Inverse reinforcement learning and variational inference, where the reward function can learn the decision maker's intentions.
The reward function is implemented as an artificial neural network that can predict the distribution of compensation through variable inference, and the results of the experiments using commercial food delivery datasets to detect riders who are abusing in sequential data proved that the proposed methods perform better than existing machine learning and deep learning methods.
Through the commercial system which the methodology is applied, we expect to have the effect of establishing a safe delivery culture with regulation of more than 100,000 food delivery riders and ensuring fair opportunities for all riders to prevent accidents from excessive competition.
SKKU's Department of Applied Data Science is a re-education type general graduate school department for corporate employees. It means that Dae-Young who published the paper in the best academic journal in the field of machine learning and information management was working while writing his dissertation.
Dae-Young said, "It took a lot of effort to write the paper while working, but I was able to achieve the results with the help of the members of the department, including the professors and staff. Also, I think I was very lucky to that the results were published in the best academic journal."
He also expressed his ambition, saying, "As a graduate of applied data science at SKKU, I would like to challenge myself to discover new research subjects at a corporation in the future."