• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.2
• /
• pp.75.1-75.1
• /
• 2019

We introduce a novel method for reconstructing the projected dark matter mass maps of merging galaxy clusters by applying the convolutional neural network (CNN) to their weak lensing maps. We generate synthesized grayscale images from given weak lensing maps that preserve their averaged galaxy ellipticity. We then apply them to multi-layered CNN with architectures of alternating convolution and trans-convolution filters to predict the mass maps. We train our architecture with 1,000 Subaru/Suprime-Cam mock weak lensing maps, and our method have better mass map prediction than the Kaiser-Squires method with the following three aspects: (1) better pixel-to-pixel correlation, (2) more accurate finding of density peak position, and (3) free from mass-sheet degeneracy. We also apply our method to the HST weak lensing map of the El Gordo cluster and compare our result to the previous studies.

• Journal of Communications and Networks
• /
• v.12 no.5
• /
• pp.433-441
• /
• 2010

Multihop transmission is a promising technique that helps in achieving broader coverage (excellent network connectivity) and preventing the impairment of wireless channels. This paper proposes a cluster-based multihop wireless network that makes use of the advantages of multihop relaying, i.e., path loss gain, and partial relay selection in each hop, i.e., spatial diversity. In this partial relay selection, the node with the maximum instantaneous channel gain will serve as the sender for the next hop. With the proposed protocol, the transmit power and spectral efficiency can be improved over those in the case of direct transmission and conventional multihop transmission. Moreover, at a high signal-to-noise ratio (SNR), the performance of the system with at least two nodes in each cluster is dependent only on the last hop and not on any of the intermediate hops. For a practically feasible decode-and-forward relay strategy, a compact expression for the probability density function of the end-to-end SNR at the destination is derived. This expression is then used to derive closed-form expressions for the outage probability, average symbol error rate, and average bit error rate for M-ary square quadrature amplitude modulation as well as to determine the spectral efficiency of the system. In addition, the probability of SNR gain over direct transmission is investigated for different environments. The mathematical analysis is verified by various simulation results for demonstrating the accuracy of the theoretical approach.

• Journal of Korea Society of Industrial Information Systems
• /
• v.17 no.7
• /
• pp.17-24
• /
• 2012

As the resources and power on Ad-hoc networks are limited, new data aggregation techniques are required for energy efficiency. The current research on data aggregation techniques is actively in progress, but existing studies don't consider the density of nodes. If nodes are densely placed in a particular area, the information which the sensor nodes placed on those areas detect can be judged as very strong association. But, the energy spent transmitting this information is a waste of energy. In this paper the densely-concentrated node area is designated as Hotspot_Zone in the multi-hop clustering environment using the AMC and a key node is selected in the area. If the request message of data aggregation is transmitted, the key node among the neighboring nodes sends its environmental information to a manager to avoid duplicate sensing information. Therefore, the life of networks can be prolonged due to this.

• Korean Journal of Human Ecology
• /
• v.13 no.6
• /
• pp.997-1005
• /
• 2004

The purpose of this research is to measure the sensibility and preference of American consumers for domestically-manufactured cotton fabrics through 7-points scale questionnaires and the 4-dimensions and 8-axes system. The key words included such images as masculine-feminine, new-old, casual-classic, and ambiguous-orderly. The images of cotton fabrics were classified through a cluster analysis. This research also included an investigation of relationship between sensibilities for cotton fabrics and their physical characteristics. The sensibilities of cotton fabrics were classified into 4 groups: 'feminine sensibility,' 'masculine sensibility,' 'new sensibility,' and 'casual sensibility.' This result represents that American consumers' sensibility is simpler than Koreans'. The order of preference was 'feminine sensibility', 'masculine sensibility', 'new sensibility', and 'casual sensibility.' The sensibilities of cotton fabrics were explained significantly by stiffness, weight, weft density, value, and chroma. Specifically, those were more influenced by color characteristics such as value, chroma, and hue than by physical characteristics. We should be able to tell a difference in the reaction of a trading country's consumers and domestic consumers to domestically-manufactured cotton fabrics through comparing this research with the previous work(Shin & Lee, 2002). Additionally, we will be able to establish a basic strategy for successfully advancing Korean fabrics into the American domestic textile industry. In this way we can expect to increase the competitive power of our domestic clothes brands.

• Korean Journal of Environment and Ecology
• /
• v.8 no.2
• /
• pp.154-159
• /
• 1995

The Chuwang valley-Kumunkwangi valley forest in Chuwangsan area was studied to investigate forest structure in relation to altitude and part of slope. Forty eight quadrats were set up in the valley forest along altitude of 470m to 780m and part of the slope. Density and basal area of trees in tree strata decreased as increasing elevation. With increasing elevation the importance values of Quercus mongolica, Fraxinus rhynchophylla increased, while those of Pinus densiflora, Lindera obtussiloba decreased. As going from lower part to upper part of the slope, the importance values of Quercus valiabilis and Lindera obtussiloba increased while those of Fraxius rhynchophylla, Acer mono decreased. Species diversity tended to decreased as going to upper parts of the slope. The range of similarity indices between elevation belts, and parts of the slope were 74.4~84.2% and 68.0~96.3%, respectively. According to importance value and cluster analysis, the studied valley forest was classified into three forest communities of Pinus densiflora-deciduous tree species community of lower part of slope, Pinus densiflora-Quercus variabilis community of middle and upper part of slope, Pinus densiflora community of the top area.

• Journal of the Korean Geographical Society
• /
• v.40 no.2 s.107
• /
• pp.187-201
• /
• 2005

With the trend of urban polynucleation, the issue of detecting economic clusters or urban employment centers has been considered as crucial. However, the prior researches had some limitations in detecting economic clusters in the empirical analysis: i.e. inherent inefficiency of density-based clustering methods, difficulty in detecting linear types of spatial clusters and lacks of consideration of economic significance. The purpose of this paper is to propose the clustering method with the procedure of testing statistical and economic significance named as VCEC (Variable Clumping method for Economic Clusters) and to apply it to a case analysis of Erie County, New York, in order to test its validity. By applying a search radius and a total employment as an economic threshold, 'the both statistically and economically significant clusters' were detected in the Erie County, and proved to be efficient.

• Journal of Internet Computing and Services
• /
• v.13 no.5
• /
• pp.1-8
• /
• 2012

In this paper, we propose a new clustering combination based on numerical data driven for rule generation mechanism. In large and complicated space, a clustering method can obtain limited performance results. To overcome the single clustering method problem, hybrid combined methods can solve problem to divided simple cluster estimation. Fundamental structure of the proposed method is combined by mountain clustering and modified Chen clustering to extract detail cluster information in complicated data distribution of non-parametric space. It has automatic rule generation ability with advanced density based operation when intelligent systems including neural networks and fuzzy inference systems can be generated by clustering results. Also, results of the mechanism will be served to information of decision support system to infer the useful knowledge. It can extend to healthcare and medical decision support system to help experts or specialists. We show and explain the usefulness of the proposed method using simulation and results.

• Architectural research
• /
• v.21 no.4
• /
• pp.91-98
• /
• 2019

In modern times, numerous cities are competing to create the unique skyline adjacent to the water. Tall buildings located across the river have a great contribution to the skyline of a riverfront city and can be a precious asset for the city. Moreover, in several cities, tall buildings and their impact on the urban skyline are a matter that should be considered and regulated in urban design. Therefore, as a prominent element in a larger visual setting of the city, tall buildings should improve the visual quality of the skyline rather than diminish that quality. This research attempts to provide an objective method to analyze the visual quality of the skyline made by a group of tall buildings through their feature of heights and spatial arrangement from riverfront views. The analysis is determined by the design variables of building heights variation, heights transition, density, and spacing of a group of tall buildings. A comparative case study of tall buildings in Yeouido and Lujiazui was conducted to prove the effectiveness of the analysis. The proposed method can be used in a simple way in the quantitative approach to quantify the visual quality of the skyline. In conclusion, Yeuido's skyline is not quite interesting from the riverfront view in terms of height variation and continuity of the skyline view because they are dispersed. Conversely, Lujiazui's skyline from the riverfront vantage points has a good quality in all aspects of the feature of height and spatial arrangements of tall buildings cluster. These factors can be used for the urban designer on how proposed tall buildings within the cluster should appropriately respond to adding image on the skyline.

• Journal of The Korean Astronomical Society
• /
• v.48 no.2
• /
• pp.155-164
• /
• 2015

In Kang (2015) we calculated the acceleration of cosmic-ray electrons at weak spherical shocks that are expected to form in the cluster outskirts, and estimated the diffuse synchrotron radiation emitted by those electrons. There we demonstrated that, at decelerating spherical shocks, the volume integrated spectra of both electrons and radiation deviate significantly from the test-particle power-laws predicted for constant planar shocks, because the shock compression ratio and the flux of inject electrons decrease in time. In this study, we consider spherical blast waves propagating through a constant density core surrounded by an isothermal halo with ρ ∝ r⁻ⁿ in order to explore how the deceleration of the shock affects the radio emission from accelerated electrons. The surface brightness profile and the volumeintegrated radio spectrum of the model shocks are calculated by assuming a ribbon-like shock surface on a spherical shell and the associated downstream region of relativistic electrons. If the postshock magnetic field strength is about 0.7 or 7 µG, at the shock age of ∼ 50 Myr, the volume-integrated radio spectrum steepens gradually with the spectral index from α_inj to α_inj + 0.5 over 0.1–10 GHz, where α_inj is the injection index at the shock position expected from the diffusive shock acceleration theory. Such gradual steepening could explain the curved radio spectrum of the radio relic in cluster A2266, which was interpreted as a broken power-law by Trasatti et al. (2015), if the relic shock is young enough so that the break frequency is around 1 GHz.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.47 no.3
• /
• pp.83-89
• /
• 2010

Scalable routing is one of the key challenges in designing and operating large scale MANETs. Performance of routing protocols proposed so far is only guaranteed under various limitation, i.e., dependent of the number of nodes in the network or needs the location information of destination node. Due to the dependency to the number of nodes in the network, as the number of nodes increases the performance of previous routing protocols degrade dramatically. We propose Cluster Overlay Dynamic Source Routing (CODSR) protocol. We conduct performance analysis by means of computer simulation under various conditions - diameter scaling and density scaling. Developed algorithm outperforms the DSR algorithm, e.g., more than 90% improvement as for the normalized routing load. Operation of CODSR is very simple and we show that the message and time complexity of CODSR is independent of the number of nodes in the network which makes CODSR highly scalable.