• Journal of Information Processing Systems
• /
• v.16 no.4
• /
• pp.809-819
• /
• 2020

Gender prediction accuracy increases as convolutional neural network (CNN) architecture evolves. This paper compares voting and ensemble schemes to utilize the already trained five CNN models to further improve gender prediction accuracy. The majority voting usually requires odd-numbered models while the proposed softmax-based voting can utilize any number of models to improve accuracy. The ensemble of CNN models combined with one more fully-connected layer requires further tuning or training of the models combined. With experiments, it is observed that the voting or ensemble of CNN models leads to further improvement of gender prediction accuracy and that especially softmax-based voters always show better gender prediction accuracy than majority voters. Also, compared with softmax-based voters, ensemble models show a slightly better or similar accuracy with added training of the combined CNN models. Softmax-based voting can be a fast and efficient way to get better accuracy without further training since the selection of the top accuracy models among available CNN pre-trained models usually leads to similar accuracy to that of the corresponding ensemble models.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.17 no.31
• /
• pp.73-80
• /
• 1994

Sequentiality in decision making is an inherent characteristic of the R&D Process, Conceptual changes are noted during the Course of the Project which represent a continuous improvement in the quality of the data available during the various project stages. In this paper, Eight characteristic types of project evaluation models have been developed economic index models, portfolio models, decision theory models, risk analysis models, frontier models, scoring models, profile models and checklists. Each of these will be critically reviewed and appraised.

• Journal of the Ergonomics Society of Korea
• /
• v.36 no.3
• /
• pp.255-265
• /
• 2017

Objective: This study aims to survey visual search performance models to assess and predict individual's visual tasks in everyday life and industrial sites. Background: Visual search is one of the most frequently performed and critical activities in everyday life and works. Visual search performance models are needed when designing or assessing the visual tasks. Method: This study was mainly based on survey of literatures related to ergonomics relevant journals and web surfing. In the survey, the keywords of visual search, visual search performance, visual search model, etc. were used. Results: On the basis of the purposes, developing methods and results of the models, this study categorized visual search performance models into six groups: probability-based models, SATO models, visual lobe-based models, computer vision models, neutral network-based models and detection time models. Major models by the categories were presented with their advantages and disadvantages. More models adopted the accuracy among two factors of accuracy and speed characterizing visual tasks as dependent variables. Conclusion: This study reviewed and summarized various visual search performance models. Application: The results would be used as a reference or tool when assessing the visual tasks.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.10 no.16
• /
• pp.149-158
• /
• 1987

The Adoption/Diffusion of Innovations(New Product), a topic of study and research that has frown rapidly in the past few decades, deals with how a new product is adopted in a society. It is of high importance to marketing organizations because New Products must be brought out continuously in order to service. The purpose of this paper is to examine the Adoption/Diffusion Models for New product which will help to analyze the Adoption/Diffusion process of Adopters. There are a number of models that, with varying degrees of success, have been used to predict market acceptance of new product. In this paper, following types of new product Adoption/Diffusion Models was suggested. (1) Adoption Models : The Alternative Models of Adoption. The Rogers Model of the Innovation Decision Process. (2) Diffusion Models : First Purchase Models(Basic Models, Extension of the Basic Models), Repeat Purchase Models

• Journal of Korean Elementary Science Education
• /
• v.28 no.4
• /
• pp.450-466
• /
• 2009

The purpose of this study was to explore preservice elementary teachers' perceptions on models used in science and science education. Participants were sixty-one undergraduate students who were enrolled in a science education course offered at a university of education located in a mid-sized city, Korea. Data were obtained from the participants at the beginning of the course when they provided their answers to a questionnaire about models. The analysis revealed that a large number of the preservice teachers perceived models as representative of physical realities. By contrast, a relatively small number of them viewed models as representations of ideas or things like theories or hypotheses. Lots of the participants were apt to define a model from the perspective of its functions and considered the purposes of models communication, teaching, and understanding as well as visualization, simplification, and clarification. Most of the preservice teachers believed that there could be multiple models for a single target, and all of them answered that models could be changed in science. It was therefore concluded that the preservice teachers perceived properly the multiplicity and variability of models. Nevertheless, they could not elaborate how a model is used and evaluated in the process of scientific inquiry, and just a few of them mentioned the detailed nature of models. The preservice teachers possessed teacher-centered views of using models in the science classroom, and a small number of them remarked that they were going to use models for students to develop their own models and perform scientific inquiry.

• The korean journal of orthodontics
• /
• v.46 no.1
• /
• pp.13-19
• /
• 2016

Objective: The aim of this study was to compare the accuracy of Bolton analysis obtained from digital models scanned with the Ortho Insight three-dimensional (3D) laser scanner system to those obtained from cone-beam computed tomography (CBCT) images and traditional plaster models. Methods: CBCT scans and plaster models were obtained from 50 patients. Plaster models were scanned using the Ortho Insight 3D laser scanner; Bolton ratios were calculated with its software. CBCT scans were imported and analyzed using AVIZO software. Plaster models were measured with a digital caliper. Data were analyzed with descriptive statistics and the intraclass correlation coefficient (ICC). Results: Anterior and overall Bolton ratios obtained by the three different modalities exhibited excellent agreement (> 0.970). The mean differences between the scanned digital models and physical models and between the CBCT images and scanned digital models for overall Bolton ratios were $0.41{\pm}0.305%$ and $0.45{\pm}0.456%$, respectively; for anterior Bolton ratios, $0.59{\pm}0.520%$ and $1.01{\pm}0.780%$, respectively. ICC results showed that intraexaminer error reliability was generally excellent (> 0.858 for all three diagnostic modalities), with < 1.45% discrepancy in the Bolton analysis. Conclusions: Laser scanned digital models are highly accurate compared to physical models and CBCT scans for assessing the spatial relationships of dental arches for orthodontic diagnosis.

• The korean journal of orthodontics
• /
• v.52 no.1
• /
• pp.20-28
• /
• 2022

Objective: Although, digital models have recently been used in orthodontic clinics, physical models are still needed for a multitude of reasons. The purpose of this study was to assess whether the printed models can replace the plaster models by evaluating their accuracy in reproducing intermaxillary relationships and by appraising the clinicians' ability to measure the printed models. Methods: Twenty sets of patients' plaster models with well-established occlusal relationships were selected. Models were scanned using an intraoral scanner (Trios 3, 3Shape Dental System) by a single operator. Printed models were made with ZMD-1000B light-curing resin using the stereolithography method 3-dimensional printer. Validity, reliability, and reproducibility were evaluated using measurements obtained by three operators. Results: In evaluation of validity, all items showed no significant differences between measurements taken from plaster and printed models. In evaluation for reliability, significant differences were found in the distance between the gingival zeniths of #23-#33 (DZL_3) for the plaster models and at #17-#43 (DZCM_1) for the printed models. In evaluation for reproducibility, the plaster models showed significant differences between operators at midline, and printed models showed significant differences at 7 measurements including #17-#47 (DZR_7). Conclusions: The validity and reliability of intermaxillary relationships as determined by the printed model were clinically acceptable, but the evaluation of reproducibility revealed significant inter-operator differences. To use printed models as substitutes for plaster models, additional studies on their accuracies in measuring intermaxillary relationship are required.

• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 2001.09a
• /
• pp.295-301
• /
• 2001

For the seismic analysis of typical roadway bridges provisions are given in most codes for analysis models, which describes however only fundamental modelling methods according to the basic theories of structural dynamics. In practice even conventional non-seismic analysis models, separate super- and substructure models, are applied, which are not adequate because of neglecting connection elements. In this study three typical roadway bridges, a Steel box bridge, a PC beam bridge and a PC box bridge are selected and simple models integrating super- and substructure as well as connection elements are given. The simple models are composed with frame elements with lumped masses representing stiffness and mass characteristics of the selected bridges. To check the properness of the simple models, analysis results with the simple models are compared with those obtained with detailed models in view of bridge failure mechanisms. It is proved that the simple models can be used in the preliminary design phase fur the determination of failure mechanisms of typical roadway bridges.

• Communications for Statistical Applications and Methods
• /
• v.18 no.2
• /
• pp.237-244
• /
• 2011

In call forecasting literature, both the seasonal autoregressive integrated moving average(ARIMA) type models and seasonal linear models have been popularly suggested as competing models. However, their parallel comparison for the forecasting accuracy was not strictly investigated before. This study evaluates the accuracy of both the seasonal linear models and the seasonal ARIMA-type models when predicting intra-day call arrival rates using both real and simulated data. The seasonal linear models outperform the seasonal ARIMA-type models in both one-day-ahead and one-week-ahead call forecasting in our empirical study.

• Water for future
• /
• v.19 no.3
• /
• pp.267-276
• /
• 1986

This study aims at the development of flood runoff model by comparing and analyzing nonlinear models with linear models in rier basins. The models which are used at the analysis are Nash model and Runoff function method as linear models, and Tank model and Storage function method as nonlinear models. The results, which are obtained from the analysis of these models by using hydrologic data of a representative basin in Nakdong river, Wi-chun basin, show that the peak time, peak flow and flood hydrogrphs by nonlinear models are better than those by linear models in comparison with observed ones, and that nonlinear models are suittable as flood runoff model.