• The Korean Journal of Applied Statistics
• /
• v.24 no.2
• /
• pp.425-433
• /
• 2011

Reject inference in credit scoring is a statistical approach to adjust for nonrandom sample bias due to rejected applicants. Function estimation approaches are based on the assumption that rejected applicants are not necessary to be included in the estimation, when the missing data mechanism is missing at random. On the other hand, the density estimation approach by using mixture models indicates that reject inference should include rejected applicants in the model. When mixture models are chosen for reject inference, it is often assumed that data follow a normal distribution. If data include missing values, an application of the normal mixture model to fully observed cases may cause another sample bias due to missing values. We extend reject inference by a multivariate normal mixture model to handle incomplete characteristic variables. A simulation study shows that inclusion of incomplete characteristic variables outperforms the function estimation approaches.

• Communications for Statistical Applications and Methods
• /
• v.12 no.1
• /
• pp.125-137
• /
• 2005

Many discriminant analysis models for binary data have been used in real applications, but none of the classification models dominates in all varying circumstances(Asparoukhov & Krzanowski(2001)). Lee and Hwang (2003) proposed a new classification model by using multinomial distribution with the maximum entropy estimation method. The model showed some promising results in case of small number of variables, but its performance was not satisfactory for large number of variables. This paper explores to use the iterative cross entropy minimization estimation method in replace of the maximum entropy estimation. Simulation experiments show that this method can compete with other well known existing classification models.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.12
• /
• pp.1-11
• /
• 1993

This paper concentrates on models useful for analyzing the error performance of ML(Maximum Likelihood) estimators of a single unknown signal parameter: that is the error intensity model. We first develop the point process representation for the estimation error and the conditional distribution of the estimator as well as the distribution of error candidate point process. Then the error intensity function is defined as the probability dessity of the estimate and the general form of the error intensity function is derived. We then develop several intensity models depending on the way we choose the candidate error locations. For each case, we compute the explicit form of the intensity function and discuss the trade-off among models as well as the extendability to the case of multiple parameter estimation.

• Communications for Statistical Applications and Methods
• /
• v.11 no.2
• /
• pp.381-397
• /
• 2004

The marginal likelihood has become an important tool for model selection in Bayesian analysis because it can be used to rank the models. We discuss the marginal likelihood for Poisson regression models that are potentially useful in small area estimation. Computation in these models is intensive and it requires an implementation of Markov chain Monte Carlo (MCMC) methods. Using importance sampling and multivariate density estimation, we demonstrate a computation of the marginal likelihood through an output analysis from an MCMC sampler.

• The Journal of Korea Robotics Society
• /
• v.14 no.2
• /
• pp.114-121
• /
• 2019

This paper presents a new benchmark system for visual odometry (VO) and monocular depth estimation (MDE). As deep learning has become a key technology in computer vision, many researchers are trying to apply deep learning to VO and MDE. Just a couple of years ago, they were independently studied in a supervised way, but now they are coupled and trained together in an unsupervised way. However, before designing fancy models and losses, we have to customize datasets to use them for training and testing. After training, the model has to be compared with the existing models, which is also a huge burden. The benchmark provides input dataset ready-to-use for VO and MDE research in 'tfrecords' format and output dataset that includes model checkpoints and inference results of the existing models. It also provides various tools for data formatting, training, and evaluation. In the experiments, the exsiting models were evaluated to verify their performances presented in the corresponding papers and we found that the evaluation result is inferior to the presented performances.

• Genomics & Informatics
• /
• v.21 no.2
• /
• pp.27.1-27.7
• /
• 2023

Recombination events complicate the evolutionary history of populations and species and have a significant impact on the inference of isolation-with-migration (IM) models. However, several existing methods have been developed, assuming no recombination within a locus and free recombination between loci. In this study, we investigated the effect of recombination on the estimation of IM models using genomic data. We conducted a simulation study to evaluate the consistency of the parameter estimators with up to 1,000 loci and analyze true gene trees to examine the sources of errors in estimating the IM model parameters. The results showed that the presence of recombination led to biased estimates of the IM model parameters, with population sizes being more overestimated and migration rates being more underestimated as the number of loci increased. The magnitude of the biases tended to increase with the recombination rates when using 100 or more loci. On the other hand, the estimation of splitting times remained consistent as the number of loci increased. In the absence of recombination, the estimators of the IM model parameters remained consistent.

• Journal of the Korean Data and Information Science Society
• /
• v.16 no.3
• /
• pp.651-656
• /
• 2005

In this paper the quasi-likelihood function was proposed and the estimators which are the solutions of the estimating equations for estimation of a class of nonlinear time series models. We compare the performances of the proposed estimators with those of the ML estimators under the heavy-railed distributions by simulation.

• Journal of Korean Society of Transportation
• /
• v.18 no.3
• /
• pp.55-76
• /
• 2000

We propose a general solution methodology for identifying the optimal aggregation interval sizes as a function of the traffic dynamics and frequency of observations for four cases : i) link travel time estimation, ii) corridor/route travel time estimation, iii) link travel time forecasting. and iv) corridor/route travel time forecasting. We first develop statistical models which define Mean Square Error (MSE) for four different cases and interpret the models from a traffic flow perspective. The emphasis is on i) the tradeoff between the Precision and bias, 2) the difference between estimation and forecasting, and 3) the implication of the correlation between links on the corridor/route travel time estimation and forecasting, We then demonstrate the Proposed models to the real-world travel time data from Houston, Texas which were collected as Part of the Automatic Vehicle Identification (AVI) system of the Houston Transtar system. The best aggregation interval sizes for the link travel time estimation and forecasting were different and the function of the traffic dynamics. For the best aggregation interval sizes for the corridor/route travel time estimation and forecasting, the covariance between links had an important effect.

• Journal of Broadcast Engineering
• /
• v.27 no.2
• /
• pp.174-184
• /
• 2022

3D human pose estimation models can be classified into a multi-view model and a single-view model. In general, the multi-view model shows superior pose estimation performance compared to the single-view model. In the case of the single-view model, the improvement of the 3D pose estimation performance requires a large amount of training data. However, it is not easy to obtain annotations for training 3D pose estimation models. To address this problem, we propose a method to generate pseudo ground-truths of multi-view human pose data from a multi-view model and exploit the resultant pseudo ground-truths to train a single-view model. In addition, we propose a multi-view consistency loss function that considers the consistency of poses estimated from multi-view images, showing that the proposed loss helps the effective training of single-view models. Experiments using Human3.6M and MPI-INF-3DHP datasets show that the proposed method is effective for training single-view 3D human pose estimation models.

• Communications for Statistical Applications and Methods
• /
• v.10 no.2
• /
• pp.585-594
• /
• 2003

It has been generally recognized that conventional binomial or Poisson model provides poor fits to the actual correlated binary data due to the extra-binomial variation. A number of generalized statistical models have been proposed to account for this additional variation. Among them, beta-binomial, correlated-binomial, and modified-binomial models are binomial-related models which are frequently used in modeling the sum of n correlated binary data. In many situations, it is reasonable to assume that n correlated binary data are exchangeable, which is a special case of correlated binary data. The sum of n exchangeable correlated binary data is modeled relatively well when the above three binomial-related models are applied. But the estimation results of correlation coefficient turn to be quite different. Hence, it is important to identify which model provides better estimates of model parameters(success probability, correlation coefficient). For this purpose, a small-scale simulation study is performed to compare the behavior of above three models.