• Journal of architectural history
• /
• v.30 no.3
• /
• pp.21-31
• /
• 2021

This paper studied the Pingzuo(平坐) platform structure of the two story building covered with one roof during the early period of Tang dynasty, based on wall paintings, stone pagodas, brick buildings and wooden buildings might be influenced by the Tang style. Instead of Chazhuzao(叉柱造), the typical column linkage in the Song, Liao and Jin buildings, it put the boundary column just behind the wall of a bracket set. Otherwise, the column root might be seen from outside, because its bracket set was still using Touxinzao(偸心造) which did not have a lateral arm on it. And its flooring structure was also different from the Song style, it used cantilever beams instead of lateral beams supported by bracket sets.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.25 no.3
• /
• pp.300-310
• /
• 2022

A vast amount of labeled data is required for deep neural network training. A typical strategy to improve the performance of a neural network given a training data set is to use data augmentation technique. The goal of this work is to offer a novel image augmentation method for improving object detection accuracy. An object in an image is removed, and a similar object from the training data set is placed in its area. An in-painting algorithm fills the space that is eliminated but not filled by a similar object. Our technique shows at most 2.32 percent improvements on mAP in our testing on a military vehicle dataset using the YOLOv4 object detector.

• Journal of KIISE:Software and Applications
• /
• v.31 no.7
• /
• pp.859-868
• /
• 2004

We propose a method of selecting initial training examples for active learning so that it can reach high accuracy faster with fewer further queries. Our method is based on the assumption that an active learner can reach higher performance when given an initial training set consisting of diverse and typical examples rather than similar and special ones. To obtain a good initial training set, we first cluster examples by using k-means clustering algorithm to find groups of similar examples. Then, a representative example, which is the closest example to the cluster's centroid, is selected from each cluster. After these representative examples are labeled by querying to the user for their categories, they can be used as initial training examples. We also suggest a method of using the centroids as initial training examples by labeling them with categories of corresponding representative examples. Experiments with various text data sets have shown that the active learner starting from the initial training set selected by our method reaches higher accuracy faster than that starting from randomly generated initial training set.

• Proceedings of the Society of Korea Industrial and System Engineering Conference
• /
• 2002.05a
• /
• pp.237-242
• /
• 2002

The sales activity of most of small manufacturing companies is based on orders of buyers. The process of promotion, receipt and selection of orders of the manufacturers is closely coupled with the load status of the production lines. The decision on whether to accept an order or not, or the selection of optimal order set among excessive orders is entirely dependent on the schedule of production lines. However, in the real world, since the production scheduling activity is mainly performed by human experts, most of small manufacturers are suffer from being unable to meet due dates, lack of rapid decision on the acceptance of new order. To cope with this problem, this paper deals with the development of an agent for selecting an optimal order set automatically. The main engine of selection agent is based on the typical job-shop scheduling model since our target domain is the injection molding company. To solve the problem, we have formulated it as IP (Integer Program) model, and it has been successfully implemented by ILOG and selection agent. And we have suggested an architecture of an agent for tackling web based order selection problems.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.25 no.5
• /
• pp.1-8
• /
• 2002

The sales activity of most of small manufacturing companies is based on orders of buyers. The process of promotion, receipt and selection of orders of the manufacturers is closely coupled with the load status of the production lines. The decision on whether to accept an order or not, or the selection of optimal order set among excessive orders is entirely dependent on the schedule of production lines. However, in the real world, since the production scheduling activity is mainly performed by human experts, most of small manufacturers are suffer from being unable to meet due dates, lack of rapid decision on the acceptance of new order. To cope with this problem, this paper deals with the development of an agent for selecting an optimal order set automatically. The main engine of selection agent is based on the typical job-shop scheduling model since our target domain is the injection molding company. To solve the problem, we have formulated it as IP (Integer Program) model, and it has been successfully implemented by ILOG and selection agent. And we have suggested an architecture of an agent for tackling web based order selection problems.

• Journal of the Korean Society for Precision Engineering
• /
• v.22 no.3 s.168
• /
• pp.68-77
• /
• 2005

The set-up of an injection molding process is a ye complicated and time-consuming job because it is required to well determine a lot of variables closely related to products. Thus, the productivity of the set-up process mainly depends on operators' expertise and know-how. To solve the problem mentioned before, this research constructs a support system which helps operators determining the condition of the injection molding easily and systematically. The construction of the support system consists of the following four steps: 1) to determine the control variables which affect the target defect types, 2) to design and implement UI(user interface) using a scenario of set-up process, 3) to design and implement the search algorithms for the initial and optima] condition, and 4) to construct the embedded system which integrates the support system with the operating system of a plastic injection molding machine. The test experiments of some typical products are performed using the embedded system to verify the validity of the support system.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2001.01a
• /
• pp.278-282
• /
• 2001

The discovery of tacit knowledge from domain experts is one of the most exciting challenges in today\`s knowledge management. The nature of decision knowledge in determining the quality a firm\`s short-term liquidity is full of abstraction, ambiguity, and incompleteness, and presents a typical tacit knowledge extraction problem. In dealing with knowledge discovery of this nature, we propose a scheme that integrates both knowledge elicitation and knowledge discovery in the knowledge engineering processes. The knowledge elicitation component applies the Verbal Protocol Analysis to establish industrial cases as the basic knowledge data set. The knowledge discovery component then applies fuzzy clustering to the data set to build a fuzzy knowledge based system, which consists of a set of fuzzy rules representing the decision knowledge, and membership functions of each decision factor for verifying linguistic expression in the rules. The experimental results confirm that the proposed scheme can effectively discover the expert\`s tacit knowledge, and works as a feedback mechanism for human experts to fine-tune the conversion processes of converting tacit knowledge into implicit knowledge.

• Journal of Information Processing Systems
• /
• v.16 no.2
• /
• pp.329-342
• /
• 2020

In recent years, emotional text classification is one of the essential research contents in the field of natural language processing. It has been widely used in the sentiment analysis of commodities like hotels, and other commentary corpus. This paper proposes an improved W-LDA (weighted latent Dirichlet allocation) topic model to improve the shortcomings of traditional LDA topic models. In the process of the topic of word sampling and its word distribution expectation calculation of the Gibbs of the W-LDA topic model. An average weighted value is adopted to avoid topic-related words from being submerged by high-frequency words, to improve the distinction of the topic. It further integrates the highest classification of the algorithm of support vector machine based on the extracted high-quality document-topic distribution and topic-word vectors. Finally, an efficient integration method is constructed for the analysis and extraction of emotional words, topic distribution calculations, and sentiment classification. Through tests on real teaching evaluation data and test set of public comment set, the results show that the method proposed in the paper has distinct advantages compared with other two typical algorithms in terms of subject differentiation, classification precision, and F1-measure.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.6 no.9
• /
• pp.2052-2072
• /
• 2012

Access control is an essential security component in protecting sensitive data and services from unauthorized access to the resources in mission-critical Cyber-Physical Systems (CPSs). CPSs are different from conventional information processing systems in such that they involve interactions between the cyber world and the physical world. Therefore, existing access control models cannot be used directly and even become disabled in an emergency situation. This paper proposes an adaptive Access Control model for Emergences (AC4E) for mission-critical CPSs. The principal aim of AC4E is to control the criticalities in these systems by executing corresponding responsive actions. AC4E not only provides the ability to control access to data and services in normal situations, but also grants the correct set of access privileges, at the correct time, to the correct set of subjects in emergency situations. It can facilitate adaptively responsive actions altering the privileges to specific subjects in a proactive manner without the need for any explicit access requests. A semiformal validation of the AC4E model is presented, with respect to responsiveness, correctness, safety, non-repudiation and concurrency, respectively. Then a case study is given to demonstrate how the AC4E model detects, responds, and controls the emergency events for a typical CPS adaptively in a proactive manner. Eventually, a wide set of simulations and performance comparisons of the proposed AC4E model are presented.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.17 no.8
• /
• pp.722-728
• /
• 2005

The present paper deals with heating energy estimation in Korean-style apartments, paying special attention to the effect of room condition settings. Two types of heating modes are considered: continuous single-zone and scheduled multi-zone. In the latter, zones during unoccupied periods remain unconditioned. Also analyzed are sensitivities in heating energy with respect to the air change rate and the set temperature. The energy use is estimated with TRNSYS 15, a dynamic load calculation program. Heating energy for the actual residential condition (1.0 ACH and $24^{\circ}C$) appears to be nearly the same as that for a typical design standard (1.5 ACH and $20^{\circ}C$). The air change rate affects heating energy as sensitive]y as the set temperature. For all the simulated cases, the scheduled multi-zone heating mode is more energy-efficient than the continuous single-zone. Heating energy depends appreciably on the shading factor. It is expected that considerable heating energy for apartment houses can be saved by employing a multi-zone mode along with appropriate control devices.