• Journal of Information Processing Systems
• /
• v.19 no.1
• /
• pp.67-79
• /
• 2023

In the image field, data augmentation refers to increasing the amount of data through an editing method such as rotating or cropping a photo. In this study, a generative adversarial network (GAN) image was created using CycleGAN, and various colors of dogs were reflected through data augmentation. In particular, dog data from the Stanford Dogs Dataset and Oxford-IIIT Pet Dataset were used, and 10 breeds of dog, corresponding to 300 images each, were selected. Subsequently, a GAN image was generated using CycleGAN, and four learning groups were established: 2,000 original photos (group I); 2,000 original photos + 1,000 GAN images (group II); 3,000 original photos (group III); and 3,000 original photos + 1,000 GAN images (group IV). The amount of data in each learning group was augmented using existing data augmentation methods such as rotating, cropping, erasing, and distorting. The augmented photo data were used to train the MobileNet_v3_Large, ResNet-152, InceptionResNet_v2, and NASNet_Large frameworks to evaluate the classification accuracy and loss. The top-3 accuracy for each deep neural network model was as follows: MobileNet_v3_Large of 86.4% (group I), 85.4% (group II), 90.4% (group III), and 89.2% (group IV); ResNet-152 of 82.4% (group I), 83.7% (group II), 84.7% (group III), and 84.9% (group IV); InceptionResNet_v2 of 90.7% (group I), 88.4% (group II), 93.3% (group III), and 93.1% (group IV); and NASNet_Large of 85% (group I), 88.1% (group II), 91.8% (group III), and 92% (group IV). The InceptionResNet_v2 model exhibited the highest image classification accuracy, and the NASNet_Large model exhibited the highest increase in the accuracy owing to data augmentation.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.22 no.3
• /
• pp.321-332
• /
• 1998

The purpose of this study is to provide basic data for more functional and more fitting clothing construction. The subjects were college women aged from 18 to 25 in June, 1997. Data were collected by anthropometry and photometry and analyzed by the characteristic of the upper body- The results are as follows: 1. After analyzing direct anthropometric data, 1 have analyzed the data by the characteristic in each somatotype, classified them and I have had 5 groups. Group 1 with middle height and standard type, group 2 with great height and standard type, group 3 with low height and slim type, group 4 with middle height and fat type, and group 5 with low height and a little fat type. 2. After analyzing indirect photometric data, 1 have analyzed them by the characteristic and I have had 4 groups. Group 1 with lean back type, group 2 with sway back type, group 3 with straight type and group 4 with bend forward type. 3. Through the mutual corresponding relation in the 5 groups classified with direct anthropometric measurement and the 4 groups classified with indirect photometric mea- surement, direct-group 1 comes the most corresponding to indirect-group 2, comes second to indirect-group 4, and comes third to indirect-group 1. Direct-group 2 comes the most corresponding to indirect-group 4, direct-group 3 comes the most corresponding to indirect- group 1, direct-group 4 comes the most corresponding to indirect-group 3, and direct-group 5 comes the most corresponding to indirect-group 1.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.7
• /
• pp.3319-3338
• /
• 2016

Cloud storage is becoming more and more popular because of its elasticity and pay-as-you-go storage service manner. In some cloud storage scenarios, the data that are stored in the cloud may be shared by a group of users. To verify the integrity of cloud data in this kind of applications, many auditing schemes for shared cloud data have been proposed. However, all of these schemes do not consider the access authorization problem for users, which makes the revoked users still able to access the shared cloud data belonging to the group. In order to deal with this problem, we propose a novel public auditing scheme for shared cloud data in this paper. Different from previous work, in our scheme, the user in a group cannot any longer access the shared cloud data belonging to this group once this user is revoked. In addition, we propose a new random masking technique to make our scheme preserve both data privacy and identity privacy. Furthermore, our scheme supports to enroll a new user in a group and revoke an old user from a group. We analyze the security of the proposed scheme and justify its performance by concrete implementations.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38B no.11
• /
• pp.934-936
• /
• 2013

A routing scheme for a group of mobile users for wireless ad-hoc networks is presented. The proposed scheme mines social activity patterns from wireless traces, and exploits social user group for efficient data routing among users based on a data publish approach. Simulation results based on real-world wireless traces show that our routing scheme reduces routing cost for a large mobile user group with a factor of 1.8 compared to a baseline counterpart.

• Proceedings of the Korean Association for Survey Research Conference
• /
• 2001.04a
• /
• pp.109.2-130
• /
• 2001

The purpose of this study is (1) to demonstrate that ignoring group differences may mislead to incorrect conclusion when analyzing correlation data (e.g., factor analysis), (2) to highlight the importance of the data analytic method considering group differences, and (3) to provide ways of incorporating group differences in data analysis. In study, 1, ignoring group difference in factor analysis may mislead to incorrect factor structure. To remedy this, z-transform method and group analysis tool in covariance structure models were suggested. In study 2, the group differences effect was illustrated by using real data (IQ test data).

• Survey Research
• /
• v.2 no.2
• /
• pp.109-130
• /
• 2001

The purpose of this study is (1) to demonstrate that ignoring group differences may mislead to incorrect conclusion when analyzing correlation data (e.g.. factor analysis). (2) to highlight the importance of the data analytic method considering group differences. and (3) to provide ways of incorporating group differences in data analysis. In study 1. ignoring group difference in factor analysis may mislead to incorrect factor structure. To remedy this. z-transform method and group analysis tool in covariance structure models were suggested. In study 2. the group differences effect was illustrated by using real data (IQ test data).

• Journal of Acupuncture Research
• /
• v.30 no.1
• /
• pp.1-11
• /
• 2013

Objectives : The 7-zone-diagnostic system is a device for predetermining bodily locations by measuring the energy of the living body. The purpose of this study is to examine correlations between gastroscopy results and zone-5 data from the 7-zone-diagnostic system. Methods : This study was carried out with data from gastroscopy procedures. It involved 115 patients who had been diagnosed with gastritis, esophagitis, or gastric ulcers. These patients were divided into three groups according to the different patterns of zone-5 factors AA, FL2 and FL1. Group A was made up of patients for whom the red bar graph of zone-5 was higher than the normal range for factor AA. Group B was made up of patients for whom the red bar graph of zone-5 was within the normal range for factor AA. Group C was made up of patients for whom the red bar graph of zone-5 was lower than the normal range for factor AA. Group D E and F were made up of same way as group A B and C for factor FL2. Group G H and I were made up of same way as group A B and C for factor FL1. After the collection of gastroscopy results and data on zone-5 from the 7-zone-diagnostic system, the data was analyzed statistically. Results : 1. Group D had a higher ratio of medication than group E, and this result was statistically significant. 2. Group D G had a higher ratio of medication than group E G and group E H, and this result was statistically significant. 3. Group G had a higher ratio of esophagitis than group H, and this result was statistically significant. 4. A Group B H had a higher ratio of esophagitis than group B G, group C G, and group C H, and this result was statistically significant. 5. The medication group had a high score on the function index than the non-medication group, and this result was statistically significant. Conclusions : This study suggests that there is a slight correlation between gastroscopy results and zone-5 data from the 7-zone-diagnostic system. The research resulted in significant data that are helpful for diagnosing digestive system problems through the use of the 7-zone-diagnostic system.

• Nuclear Engineering and Technology
• /
• v.46 no.3
• /
• pp.353-362
• /
• 2014

The effects of nuclear data uncertainties are studied on a typical PWR fuel assembly model in the framework of the OECD Nuclear Energy Agency UAM (Uncertainty Analysis in Modeling) expert working group. The "Fast Total Monte Carlo" method is applied on a model for the Monte Carlo transport and burnup code SERPENT. Uncertainties on $k_{\infty}$, reaction rates, two-group cross sections, inventory and local pin power density during burnup are obtained, due to transport cross sections for the actinides and fission products, fission yields and thermal scattering data.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.6 no.4
• /
• pp.293-298
• /
• 2006

Thermopiezoelectric method, using poly silicon heater and a piezoelectric sensor, was proposed for writing and reading in a probe based data storage system. Resistively heated tip writes data bits while scanning over a polymer media and piezoelectric sensor reads data bits from the self-generated charges induced by the deflection of the cantilever. 34${\times}$34 array of thermopiezoelectric nitride cantilevers were fabricated by a single step wafer level transfer method. We analyzed the noise level of the charge amplifier and measured the noise signal. With the sensor and the charge amplifier 20mn of deflection could be detected at a frequency of 10KHz. Reading signal was obtained from the cantilever array and the sensitivity was calculated.

• Asia pacific journal of information systems
• /
• v.9 no.1
• /
• pp.143-163
• /
• 1999

This study investigates the effect of group size on group performance, here the quality of group decision, Four effects are proposed and tested in a field experimental setting : (1) the relationship between the group size and the distribution of individual's problem-solving ability ; (2) the change of the group decision quality as group size increases ; (3) the relationship between the group decision quality and the quality of the best/worst member as group size increases ; (4) the relationship between the group decision quality and the average quality of individuals in the group as group size increases. Data showed that contrary to the exiting results, group decision quality was not improved with the group size. Rather, it showed a little tendency that group decision quality was worsened with the group size. Data also showed that consensus-oriented group decision making process produced the compromised output. Thus, group decision quality was not better than the average group members'. The opinion of the best member was not accepted. The implications of the findings are discussed for the development of a Korean GSS.