• The Journal of the Korea Contents Association
• /
• v.11 no.5
• /
• pp.291-301
• /
• 2011

This research describes the comparison broker's role and its effectiveness measurement framework of comparison shopping. We verified seller-led comparison challenge method provide comparison information of products to buyers more efficiently. Buyer's satisfaction of purchase (S) can be defined as an interactive function between seller's competitiveness vector (P) of products that supplied to the market, and buyer's informed level vector (B) of products that is known from a lot of sources. Then the buyer's informed level can be changed through the information analysis among products by transformation process using comparison matrix (C). So the role of comparison shopping is to construct a comparison matrix and to serve it to the buyers, and to change the buyer's informed level. The changed informed level influences a buyer's satisfaction, that improved satisfaction of purchase is defined as the effectiveness of comparison shopping. As a perfect construction and usage of comparison matrix is impossible, a more efficient method for improving the comparison effectiveness is the comparison challenge. This research shows that comparison shopping makes 9.32% and comparison challenge makes 19.11% enhancement of comparison effectiveness through television market data experiments.

• The Korean Journal of Applied Statistics
• /
• v.13 no.2
• /
• pp.371-381
• /
• 2000

In this paper, we describes smoothing spline in nonparametric regression and some asymptotic results for estimates of the regression cofficients in the parametric part were biased on semi parametric regression estimator.

• Proceedings of the KOR-KST Conference
• /
• 1998.10a
• /
• pp.226-235
• /
• 1998

본 논문에서는 추종 모형을 이용한 미시 교통류 시뮬레이션 모형을 개발하고, 이 모형을 통한 시뮬레이션을 통하여 단속류에서 검지기의 설치 위치에 따른 검지 특성을 알아보고, 교통정보 수집용의 검지기의최적 위치에 대해 평가하였다. 검지기로부터 발생하는 교통량, 점유율, 속도 자료중 링크의 통행시간을 가장 잘 반영하는 것은 점유율에 의한 검지기의최적위치는 정지선으로부터 150∼250m이다. 점유율 다음으로 통행시간을 잘 반영하는 자료는 지점속도로서 점유율보다는 상관관계가 낮지만, 양호한 설명력을 가지는 것으로 보인다. 교통량 자료는 상관관계가 낮으며, 교통량에 의한 위치 선정은 각 모의 실험 결과에서 일관적이지 않아 적절한 설명변수가 아니라고 판단하였다. 모든 경우에서, 정지선이나 링크 최상류에 위치한 검지기로부터의 자료는 통행시간과 독립적이므로 이러한 검지기는 교통정보 수집용을 사용할 수 없으며, 일반적인 검지기의최적 위치는 정상상태의 교통류 뿐만 아니라 대기행렬내에 존재하여 매우 혼잡한 상태를 경험할 수 있는 위치라고 할수 있다.

• The Korean Journal of Applied Statistics
• /
• v.1 no.1
• /
• pp.39-44
• /
• 1987

For obtaining the m.l.e. of $\Sigma$ from p-variate Non-singular Normal parent, $N_p(\mu, \Sigma)$, Andersous' Procedure based on the invariance property of the m.l.e. seems to be generally Preferred in the view of its simplicity. This paper shows that his approach with respect to $\Sigma^{-1}$ rather than $\Sigma$ itself, he burther applicable to deriving the m.l.e. of parametersinvolved in the common factor model an dsimplex model as well.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.10 no.2
• /
• pp.22-34
• /
• 2011

This paper presents the investigation results of a research on how engineers can analyze the economic effect of the ETCS(Electronic Toll Collection System) installed to minimize the vehicle delays on freeway tollgates during toll payments. This research considered this economic effect to occur in the form of vehicle passing time reductions at the ETCS, and the multi-service queuing theory was applied to estimate these values. This research found: 1) When vehicles approaching tollgates show Poisson distribution and the service time of the ETCS shows Exponential distribution, the multi-service queuing theory would be applicable for estimating vehicle passing times at toll-gates, 2) Despite the ETCS placement, exit sections of tollgates give a greater reduction of vehicle passing times than entering sections due to more delays at conventional toll payments, and 3)The ETCS would not guarantee vehicle passing time reductions all the time, because in such a case as many vehicles were queuing at the ETCS, the total delay level for a toll gate would increase greatly. In addition, in order to examine the accuracy of the estimated vehicle passing values, this research compared the values from the multi-service queuing theory with the observed values from a set of field survey values at freeway toll-gates, and found that the two values were in a good agreement with a very low error range of 1-3 seconds per vehicle. Based on this result, the multi-service queuing theory was recommended for practice.

• Journal of Korea Water Resources Association
• /
• v.32 no.6
• /
• pp.657-664
• /
• 1999

This paper introduced the basic theory of LRCS(Linear Reservoir and Channel System) rainfall runoff model proposed by Korean researchers(Lee and Lee, 1995), and discussed the change of model output according to objective functions in sensitivity analysis and calibration process of model. It proposed "hat" matrix and affluence measures for affluence analysis of parameters in calibration, and investigated relationship between change of model output according to error propagation in parameter estimation, and sensitivity of model output according to variance of model output and change of parameters. Accuracy of parameter estimates was known by analysis of sensitivity coefficient, diagonal element $h_i$ and $D_i$._i$.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.170-170
• /
• 2021

본 연구에서는 토지이용변화 예측 모형으로 산출된 토지이용도를 사용하여 용담댐 소유역의 지표유출량을 비교 및 분석하였다. 토지이용예측모형은 DynaCLUE 모형을 사용하였으며, 토지이용 면적 시나리오는 2000년, 2007년 및 2013년 실제 중분류 토지이용도를 기반으로 회귀식을 산정하였다. 모의된 토지이용도는 실제 토지이용도와 공간적인 분포 및 면적 비교를 통해 변환 탄성계수와 변환 행렬을 수정하여 검·보정하였다. DynaCLUE 모형으로 모의된 토지이용도는 공간적인 분포에서 초지가 실제 토지이용도와 차이가 발생하였으나, 각 토지이용별 면적을 비교한 경우 모의 토지이용도와 실제 토지이용도가 매우 유사하게 나타났다. CLUE 모형으로 모의된 토지이용도에서 발생하는 공간적인 불확실성은 복잡한 용담댐 소유역의 토지이용을 반영할 Driving factor가 부족하여 발생하는 것으로 판단된다. 산출된 모의 토지이용도를 SWAT 모형의 입력 자료로 사용하여 2013년 용담댐의 소유역 지표유출량을 모의하였다. SWAT으로 산정된 유출량의 보정은 SWAT-CUP의 SUFI-2 알고리즘을 이용했으며, 보정된 모의 지표유출량과 실제 유량 측정값을 비교한 결과 유의미한 비교 결과가 나타났다. 향후 토지이용예측모형을 이용하여 토지이용 변화를 수문 분석에 반영하는 추가 연구가 필요할 것으로 판단된다.

• The Korean Journal of Applied Statistics
• /
• v.36 no.6
• /
• pp.529-545
• /
• 2023

This paper introduces the Banded-VHAR model suitable for high-dimensional long-memory time series with band structure. The Banded-VHAR model has nonignorable correlations only with adjacent dimensions due to data features, for example, geographical information. Row-wise estimation method is adapted for fast computation. Also, two estimation methods, namely BIC and ratio methods, are proposed to estimate the width of band. We demonstrate asymptotic consistency of our proposed estimation methods through simulation study. Real data applications to pm2.5 and apartment trading volume substantiate that our Banded-VHAR model outperforms traditional sparse VHAR model in forecasting and easy to interpret model coefficients.

• Journal of Korean Society of Forest Science
• /
• v.104 no.1
• /
• pp.111-116
• /
• 2015

The purpose of this study was to evaluate mixed model of dbh-height relation containing random effect. Data were obtained from a survey site for Quercus variabilis in Gangwon region and remeasured the same site after three years. The mixed model were used to fixed effect in the dbh-height relation for Quercus variabilis, with random effect representing correlation of survey period were obtained. To verify the evaluation of the model for random effect, the akaike information criterion (abbreviated as, AIC) was used to calculate the variance-covariance matrix, and residual of repeated data. The estimated variance-covariance matrix, and residual were -0.0291, 0.1007, respectively. The model with random effect (AIC = -215.5) has low AIC value, comparison with model with fixed effect (AIC = -154.4). It is for this reason that random effect associated with categorical data is used in the data fitting process, the model can be calibrated to fit repeated site by obtaining measurements. Therefore, the results of this study could be useful method for developing model using repeated measurement.

• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.249-263
• /
• 2023

Collaborative Filtering, a representative recommendation system methodology, consists of two approaches: neighbor methods and latent factor models. Among these, the latent factor model using matrix factorization decomposes the user-item interaction matrix into two lower-dimensional rectangular matrices, predicting the item's rating through the product of these matrices. Due to the factor vectors inferred from rating patterns capturing user and item characteristics, this method is superior in scalability, accuracy, and flexibility compared to neighbor-based methods. However, it has a fundamental drawback: the need to reflect the diversity of preferences of different individuals for items with no ratings. This limitation leads to repetitive and inaccurate recommendations. The Adaptive Deep Latent Factor Model (ADLFM) was developed to address this issue. This model adaptively learns the preferences for each item by using the item description, which provides a detailed summary and explanation of the item. ADLFM takes in item description as input, calculates latent vectors of the user and item, and presents a method that can reflect personal diversity using an attention score. However, due to the requirement of a dataset that includes item descriptions, the domain that can apply ADLFM is limited, resulting in generalization limitations. This study proposes a Generalized Adaptive Deep Latent Factor Recommendation Model, G-ADLFRM, to improve the limitations of ADLFM. Firstly, we use item ID, commonly used in recommendation systems, as input instead of the item description. Additionally, we apply improved deep learning model structures such as Self-Attention, Multi-head Attention, and Multi-Conv1D. We conducted experiments on various datasets with input and model structure changes. The results showed that when only the input was changed, MAE increased slightly compared to ADLFM due to accompanying information loss, resulting in decreased recommendation performance. However, the average learning speed per epoch significantly improved as the amount of information to be processed decreased. When both the input and the model structure were changed, the best-performing Multi-Conv1d structure showed similar performance to ADLFM, sufficiently counteracting the information loss caused by the input change. We conclude that G-ADLFRM is a new, lightweight, and generalizable model that maintains the performance of the existing ADLFM while enabling fast learning and inference.