Computational Study of Lifted Inequalities for 0-1 Mixed-Integer Bilinear Covering Sets

Kwanghun Chung(정광헌) 한국경영과학회 2014 한국경영과학회 학술대회논문집 Vol.2014 No.5

Computational Study of Cutting Planes for a Lot-Sizing Problem in Branch-and-Cut Algorithm

Kwanghun Chung(정광헌) 한국경영과학회 2015 한국경영과학회지 Vol.40 No.3

In this paper, we evaluate the strength of three families of cutting planes for a lot-sizing problem. Lot-sizing problem is very basic MIP model for production planning and many strong valid inequalities have been developed for a variety of relaxations in the literature. To use three families of cutting planes in Branch-and-Cut framework, we develop separation algorithms for each cut and implement them in CPLEX. Then, we perform computational study to compare the effectiveness of three cuts for randomly generated instances of the lot-sizing problem.

Lifting 기법을 이용한 Generalized Bilinear Cover Inequality

정광헌(Kwanghun Chung) 한국경영과학회 2017 韓國經營科學會誌 Vol.42 No.3

In this paper, we generalize lifted inequalities to a 0-1 mixed-integer bilinear covering set with linear terms. This work is motivated by the observation that Generalized Bilinear Inequality (GBI) occurs in the Branch and Bound process. We find some conditions and prove the subadditivity of lifting functions for lifting to be sequenceindependent. Using the theoretical results, we develop facet-defining inequalities for a GBI-defined set through three steps of lifting.

Computational Study of Lifted Inequalities for 0-1 Mixed-Integer Bilinear Covering Sets

Kwanghun Chung(정광헌) 대한산업공학회 2014 대한산업공학회 춘계학술대회논문집 Vol.2014 No.5

외래 환자의 임상특성이 예약 부도에 미치는 영향 분석

이상복,박기택,정광헌,Lee, Sangbok,Park, Kitaek,Chung, Kwanghun 한국전자거래학회 2017 한국전자거래학회지 Vol.22 No.4

본 논문에서는 환자들의 임상특성이 예약 부도에 미치는 영향을 분석하고자 한다. 미국의 한 보훈병원에 있는 7,055명 환자들의 진단 상병명과 그들의 예약 부도 데이터를 이용하여 진단 상병에 따른 예약 부도의 차이, 진단 상병의 개수에 따른 차이에 대해서 분석 및 비교를 실시하였다. 약물 중독, 우울증과 같은 정신질환과, 고혈압과 같은 만성 질환에서 예약 부도가 유의하게 높게 나타났다. 진단 상병의 개수가 증가할수록 예약 부도가 감소하는 것을 확인할 수 있으나, 진단 상병의 개수가 4개를 넘어서면 예약 부도가 통계적으로 유의하게 감소하지 않는 것으로 나타났다. 이러한 통계적 분석 과정을 통해 의료 환경에서 수행하고 있는 진료 및 예약 부도 관리의 문제점을 식별하여 제시하였다. 본 논문의 결과는 병원의 예약 부도에 대한 해결책을 찾는데 이용될 수 있다고 판단된다. This study focuses on analyzing no-shows associated with patients' clinical characteristics described by diagnoses in their medical data. A dataset of 7,055 patient-records from a Veterans hospital in the United States was used to test if there is difference on no-shows along with each patient's diagnosed diseases and the number of diagnoses. Patients with mental diseases such as drug dependence abuse and major depression, and chronic diseases such as hypertension are more likely to no-show. In comparisons with the number of diagnoses, the no-show decreases as the number of diagnoses increases up to four and doesn't change significantly afterwards. We provide managerial insights on clinical operations problems from statistical analysis. We believe that our results can be used to develop appropriate solutions on no-shows in clinics.

확률적 동적 계획법을 이용한 공공자전거 정비소에서의 최적 재고관리

임희종(Heejong Lim),정광헌(Kwanghun Chung) 한국생산관리학회 2020 한국생산관리학회지 Vol.31 No.1

This study develops an optimal inventory management policy for shared bicycles that need to be prepared at the maintenance center. We consider a stochastic dynamic programming model with both the operation cost and the social cost caused by carbon emissions. If a shared bicycle needs maintenance, it is repaired at the workshop and a spare one is delivered to the bicycle rental station. If there are too many bicycle inventories, there will be unnecessary overage cost such as investment and insurance costs for bicycles. If too few spare bicycles are kept on inventory, there will not be sufficient bicycles to be placed in the rental station and this will incur underage cost. We estimate the probability distribution of repair demands caused by failure using a negative binomial distribution. Our estimated demand probability distribution is applied to the stochastic dynamic programming to derive an optimal inventory policy. 본 연구에서는 공공자전거의 운영비용 및 탄소배출량에 따른 사회적 비용과 자전거의 수리 수요 발생을 고려하여, 정비센터에서 보유해야 할 여분의 공공자전거에 대한 최적 재고관리 방안을 개발하고자 한다. 공공자전거의 수리가 필요한 경우, 고장난 자전거는 정비소에서 수리를 받고, 여분의 공공자전거는 자전거 대여소로 이송된다. 만일 정비센터에서 자전거 재고를 너무 많이 유지한다면, 자전거에 대한 불필요한 투자비용과 보험료 등이 발생하게 되며, 너무 적은 자전거 재고를 유지한다면 고장난 자전거의 수리가 필요한 경우 대여소에 배치할 자전거가 부족하여 사용자 만족도 저하 및 사회적 비용이 발생하게 된다. 본 연구에서는 공공자전거의 고장수리 요구를 수요로 고려하여, 고장에 의해 발생하는 수리 수요의 확률분포를 음이항 분포로 추정하였다. 또한, 추정된 수요의 확률분포를 확률적 동적계획법에 적용하여, 최적 재고 운영방안을 도출하였다.

상향식-하향식 DEA 모형을 이용한 온실가스 감축량 배분

민대기(Daiki Min),정광헌(Kwanghun Chung) 한국경영과학회 2018 經營 科學 Vol.35 No.3

In this paper, we develop a model that adjusts the GHG (Green House Gas) reduction targets by taking account of the bottom-up allocation approach in addition to the top-down allocations. Using environmental efficiency instead of the amount of GHG emissions, we propose a DEA (Data Envelopment Analysis)-based model to allocate the reduction target. By comparing the top-down allocations with the bottom-up allocations, we identify how much each company can reduce or should require the amount of GHG emission permits. In this regard, the purpose of this research includes the development of mathematical models, collecting data sets for empirical analysis, and showing managerial implications. The top-down model allocates the GHG reductions into companies while maximizing the total efficiency so as to accomplish fair allocations. The bottom-up model determines the maximum amount of GHG reductions with aims to maximize the efficiency of individual company. We perform empirical analysis to see differences between the top-down and bottom-up allocations. We obtain data from 36 Korean domestic companies in steel industry. Our empirical analysis provides the guideline how much each company should buy or can sell GHG emissions.

용량 할당 의사결정에서 발생하는 행동적 편향에 관한 실험적인 연구

김승범 ( Seungbeom Kim ),정광헌 ( Kwanghun Chung ) 한국로지스틱스학회 2021 로지스틱스연구 Vol.29 No.2

This study investigates whether there exists behavioral bias in the decisions of capacity allocation by employees of air freight companies. In addition, it also examines whether the act of reporting can mitigate the bias. There are two types of demands on air cargo companies: long-term and short-term. Long-term demand is confirmed in the earlier period before the departure from a cheaper price. Therefore, it can be seen as a product with low profitability but low uncertainty. Short-term demand comes later, near the time of the departure, with a high price. Short-term demand, as opposed to long-term demand, can be said to be a product with high yield but high uncertainty. This problem of appropriate allocation for short- and long-term cargo is one of the traditional problems of revenue management, and a number of theoretical studies have been conducted on this. However, to the best of our knowledge, there are very few behavioral studies that focus on the decision makers of air cargo firms actually allocating the space resource. Three experiments were conducted using Amazon Mechanical Turk. We found that there is an pull-to-center bias to the half of the capacity, which resulted in inefficiency of the allocation. In order to alleviate this behavioral bias, self-reporting was requested to the subject, and we found that the bias could be alleviated.

죽음의 계곡 벤처기업 비즈니스 프로세스 개발

황은석(Eunseok Hwang),석혜성(Hyesung Seok),정광헌(Kwanghun Chung) 한국산학기술학회 2019 한국산학기술학회논문지 Vol.20 No.3

2017년 청년 실업률이 역대 최고치를 기록한 가운데, 본 상황을 타파하기 위한 대안으로 창업 활성화가 떠오르고 있다. 벤처 기업 수 및 신규 벤처 투자 금액은 매년 역대 치를 갱신하고 있으며 정부에서는 중소기업청을 중소벤처기업부로 격상시켰다. 문제는 스타트업의 성공률이 높지 않다는 데 있다. 한국 신생기업 생존율은 최근 3년 39.1%에 불과하며, 죽음의 계곡(Valley of Death) 단계에서 90%에 달하는 기업들이 실패하고 있다. 초기 스타트업은 죽음의 계곡을 넘어갈 수 있느냐가 관건이며 이를 위해투자가 어느 정도 규모와 적시성을 가지고 이루어지느냐가 중요한 변수로작용한다. 하지만, 현재 스타트업의 비즈니스 모델을 정립하기 위한 모델들은 투자 시 실질적인 기업 평가 변수가 반영되어 있지 않아 투자 유치에 어려움이 있다. 본 논문에서는 주요 투자 주체인 엔젤투자기관 및 벤처캐피탈이 투자를 위해 중점적으로 고려하는 변수 및 비즈니스 모델이 혼합된 프로세스를 개발했다. 또한, 새로 개발한 Death Valley Venture(DVV) 비즈니스 프로세스를 기존의 기법들과 비교하고, 적용 사례를 제시하였다. Youth unemployment reached a record high in 2017, and business revitalization is emerging as a means of overcoming this situation. The number of venture companies and amounts of new venture investments are increasing year on year, and the government has upgraded Small and Medium Business Administration to the Ministry of SMEs and Startups. However, the success rate of startups is relatively poor. Over the past three years only 39.1% of Korean startups survived and 90% of companies failed in the Valley of Death phase. Survival this phase is critical for early startups, and thus, the amount and timing of investments are critical. Current models for establishing business startups do not effectively attract investments because they lack practical, corporate evaluation variables. In this paper, we develop a mixed process that incorporates the factors and business models focused on by venture capital investors. In addition, we compared our Death Valley Venture (DVV) process with current methods and provide an example of its application.

서울시 공공자전거 시스템 운영을 위한 효율적 관리 구역 설정

이상복(Sangbok Lee),임희종(Heejong Lim),정광헌(Kwanghun Chung) 한국경영과학회 2018 經營 科學 Vol.35 No.1

Like many other cities around the world, Seoul has been operating a bike sharing system since 2015, which is aiming at reducing traffic and pollution at the same time. As the use of shared bicycles increases, managerial issues such as relocation and maintenance of bicycles have been raised to offer better service to the users. In this paper, we develop a mixed-integer programming model for finding optimal clusters of bike stations for the relocation of bikes. From each bike station, net demand is obtained from user pickup and return data at 210 stations in Seoul. Then, we perform computational experiments with historical data to propose better clustering for efficient management of bike sharing systems. Computational results show that optimal clusters reduce the workload deviation among clusters.