• Journal of Intelligence and Information Systems
• /
• v.12 no.2
• /
• pp.167-183
• /
• 2006

Since data mining attempts to find unknown facts or rules by dealing with also vaguely-known data sets, it always suffers from high error rate. In order to reduce the error rate, many researchers have employed multiple models in solving a problem. In this research, we present a new type of multiple models, called DyMoS, whose unique feature is that it classifies the input data and applies the different model developed appropriately for each class of data. In order to evaluate the performance of DyMoS, we applied it to a real customer churn problem of an automobile insurance company, The result shows that the DyMoS outperformed any model which employed only one data mining technique such as artificial neural network, decision tree and case-based reasoning.

• Journal of Intelligence and Information Systems
• /
• v.12 no.4
• /
• pp.109-126
• /
• 2006

An effective customer classification has been essential for the successful customer relationship management. The typical customer rating is carried out by the proportionally allocating the customers into classes in terms of their life time values. However, since this method does not accurately reflect the homogeneity within a class along with the heterogeneity between classes, there would be many problems incurred due to the misclassification. This paper suggests a new method of rating customer using Holistic profit technique, and validates the new method using the customer data provided by an insurance company. Holistic profit is one of the methods used for deciding the cutoff score in screening the loan application. By rating customers using the proposed techniques, insurance companies could effectively perform customer relationship management and diverse marketing activities.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.42 no.2
• /
• pp.69-77
• /
• 2019

As information technology advances, the penetration of smart devices connected to the Internet, such as smart phone and tablet PC, has rapidly expanded, and as sensor prices have fallen the Internet of Things has begun to be introduced in the industry. Today's industry is rapidly changing and evolving, requiring companies to respond to the new paradigm of business. In this situation, companies need to actively manage and maintain customer relationships in order to acquire loyal customers who bring them a high return. The purpose of this study is to suggest a method to manage customer relationship using real time IoT data including IoT product usage data, customer characteristics and transaction data. This study proposes a method of segmenting customers through RFM analysis and transition index analysis. In addition, a real-time monitoring through control charts is used to identify abnormalities in product use and suggest ways of differentiating marketing for each group. In the study, 44 samples were classified as 9 churn customers, 10 potential customers, and 25 active customers. This study suggested ways to induce active customers by providing after-sales benefit for product reuse to a group of churn customers and to promote the advantages or necessity of using the product by setting the goal of increasing the frequency of use to a group of potential customers. Finally, since the active customer group is a loyal customer, this study proposed an one-on-one marketing to improve product satisfaction.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2009.04a
• /
• pp.119-124
• /
• 2009

This paper proposes a tuned liquid column sloshing damper(TLCSD) and presents experimental results to evaluate its control performance. The proposed damper acts as a tuned liquid column damper(TLCD) and a tuned liquid damper(TLD), respectively, in both principal axes of building structures. Shaking table test was performed to grasp its dynamic characteristics. Testing results showed that under inclined incident excitations, a TLCSD used in this study have dynamic characteristics coupled by both TLCD and TLD.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2016.07a
• /
• pp.45-48
• /
• 2016

오늘날의 고객은 다양한 정보를 통해 넓은 선택의 기회를 가진다. 이러한 상황에서 기업들은 고객과의 지속적인 관계를 유지하기 어려워짐에 따라 고객 유지와 신규 고객 유치를 위한 마케팅 비용을 천문학적으로 지출하고 있다. 기업들이 이탈하는 고객의 속성을 분석하고 이탈 시점을 예측할 수 있다면 마케팅에 사용되는 비용과 노력을 최소화할 수 있을 것으로 예측된다. 이를 위해 본 논문에서는 효과적인 고객 이탈 예측을 위한 딥러닝 기반의 이탈 예측 모델을 제안한다. 이 모델은 모바일 RPG 게임 고객의 시계열적인 행동 패턴을 이용하여 이탈을 예측하는 모델로, 예측을 위한 학습을 할 때 모델링된 고객 데이터를 분석하여 이탈 고객의 특성을 파악할 수 있게 한다. 실험을 통해 이탈 고객과 미 이탈 고객의 모델링된 값이 각각 특정 속성에 치중되어 있는 것을 확인하였고, 제안 모델이 합리적으로 고객의 이탈을 예측하는 것을 보였다.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2007.11a
• /
• pp.195-200
• /
• 2007

System identification of real-scale structure is performed using forced vibration test. There exist various techniques available for identifying the dynamic characteristis of structures using dynamic and static measurements. In this study, The finite element(FE) model of the structure is analytically constructed using ANSYS and the model was updated using the results experimentally measured by the forced vibration test. forced vibration tests showed that Hybrid Mass Damper induced floor responses coincided with the earthquake induced ones which was numerically calculated based on the updated FE model.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2007.11a
• /
• pp.648-654
• /
• 2007

This paper presents the result of a comparison study to evaluate the performance of several semi-active control algorithms for use with large-scale MR damper applied to a building structure under seismic excitation using real-time hybrid test method. Recently, a variety of semi-active control algorithm studies are developed and generally evaluated the performance by using numerical analysis. In this paper real-time hybrid test method was applied to performance evaluating of semi-active control algorithms including a clipped optimal algorithm and the modulated homogeneous friction algorithm.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1996.11a
• /
• pp.254-259
• /
• 1996

Phase Doppler particle analyze. (PDPA) is a instrument which can be used to obtain simultaneous size and velocity measurements in a multiphase flow. In this study, the size of the water droplets entrained from a bubbling surface of a stagnant liquid column is measured by PDPA with a specially designed transmitter of long focal length and large beam diameter. The test section tube is made of acryle with 18 mm I.D. and 900 mm length. The experimental data are obtained for the air superficial velocity between 0.7 m/s to 3.4 m/s at atmospheric pressure. The experimental results show that there exists large difference in the entrainment mechanism between the churn-turbulent flow and annular flow. Through the present study, the phase Doppler analyzer system is shown to be successfully applied to measure particle sizes larger than $2,000\mu\textrm{m}$ if a transmitter of long focal length is utilized.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1996.05b
• /
• pp.331-336
• /
• 1996

The effects of a mixing vane on air-water mixed flow have been experimentally studied in this work, to investigate the basic mechanisms that the mixing vane affects critical heat flux (CHF). Experiment was performed for various flow rates focusing on bubbly flow and annular flow patterns. Acrylic tube (1.7m long, 11 mm I.D.) and the split vane type mixing vane were used, and ring-type conductance probes were used to measure the liquid film thickness in annular flow. Experimental results show that, (a) bubbly-to slug flow transition and churn-to-annular flow transition occur respectively near the mixing vane compared to the tests without mixing vane, (b) in bubbly flow region, the mixing vane breaks the bubbles into smaller ones and forwards bubbles to the center region of the tube by the centrifugal force, (c) the liquid film thickness in annular flow is decreased near the mixing vane for mass fluxes.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2004.05a
• /
• pp.523-526
• /
• 2004

가입자단에서의 병목현상 해결과 통신방송융합서비스의 제공을 위해 가입자망 기술은 디지털화된 HFC와 상하향 100Mbps이상의 속도를 제공하는 FTTH(Fiber To The Home)로 진화될 것으로 예측되며, 궁극적으로는 FTTH로 수렴될 것으로 전망되고 있다. FTTH는 통신과 방송이 융합된 다양한 신규서비스로 새로운 수익을 창출할 것으로 기대되나, 대규모 투자가 필요한 상황에서 정확한 수요 예측은 통신 회사들의 네트워크 구축 관련 의사결정에 필수적인 요인이다. 이를 위해 본 논문에서는 경쟁서비스로 천이가 용이한 통신서비스의 특징을 고려하여 FTTH의 수요예측을 실시하였다. 1차 전문가 설문을 통해 서비스간 전이형태를 모형화하였으며, 이를 토대로 FTTH 서비스로의 천이에 의해 추가적인 확산이 일어나는 변형된 Bass모형을 제안하였다. FTTH 서비스의 확산요인을 분석한 후 이를 토대로 2차 전문가 설문을 통해 각 계수를 추정하였다.