SIAI Yearbook – 2024
Last year, the SIAI Yearbook 2023 focused on “The Use of AI Algorithms in Corporate Management”. This year, in response to the growing interest in the AI industry within the machinery sector, the Yearbook has been prepared with a focus on system efficiency.
As the use of wearable devices like Galaxy Watch and Apple Watch has been increasing, Researcher Yeonsook Kwak discusses a recent paper on biometric behavior data applied in the healthcare industry. According to Kwak, sleep tracking with wearable devices still faces challenges, especially in accurately detecting when a person falls asleep and wakes up. In the recent study, Kwak improved accuracy by using the differences in the distribution functions between sleep and awake data. Unlike simple AI algorithms that use only one-dimensional averages, distribution functions can identify complex changes by analyzing data in multiple dimensions. This approach is expected to be effective not just for sleep tracking but also for examining various other types of data collected from wearable devices.
Researcher Sungsu Han introduces a study on improving the efficiency of managing Seoul’s public bike-sharing service, “Seoul Bike”, which was launched by the city of Seoul in 2014. This study draws inspiration from major cities like New York and London, where instead of managing all bicycle parking zones, efforts are concentrated in specific areas, dividing the management zones into roughly 5 to 10 sections. According to Han, indeed, dividing Seoul’s public bicycle management zones into five large or ten small clusters would maximize efficiency. The study also addresses the inefficiency of the current system, where Seoul’s 25 districts operate independently. Han applied time-series modeling to remove the effects of the COVID-19 pandemic, as well as other seasonal, trend, and weather influences on the data. To better understand how Seoul Bike users cluster together, Han used the Louvain algorithm, which takes network effects into account. Han suggests that by implementing these methods, Seoul could save costs through better allocation of operational personnel.
Researcher Donggyu Kim examined the supply-demand balance of the blood supply chain managed by the Korean Red Cross. Despite using advanced time-series analysis techniques such as the Seasonal AutoRegressive Integrated Moving Average (SARIMA) model, which considers various variables including the COVID-19 pandemic period, weather, gender differences, and regional variations, Kim stated that the Korean Red Cross’s blood management is stable enough to prevent shortages.
Researcher HyoungKeun Kwon analyzed tax policies in overlapping jurisdictions of school districts and local districts in New York State using game theory, evaluating the utility enjoyed by residents in each area. Kwon explained that the utility of residents, as discussed in the research paper, can be translated into real-world factors such as real estate prices and living conditions. Therefore, the study could provide insights into how budget allocations for elementary, middle, and high schools by school districts might impact local communities and real estate prices.
