• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.11
• /
• pp.435-442
• /
• 2018

Outlier detection means to detect data samples that deviate significantly from the distribution of normal data. Most outlier detection methods calculate an outlier score that indicates the extent to which a data sample is out of normal state and determine it to be an outlier when its outlier score is above a given threshold. However, since the range of an outlier score is different for each data and the outliers exist at a smaller ratio than the normal data, it is very difficult to determine the threshold value for an outlier score. Further, in an actual situation, it is not easy to acquire data including a sufficient amount of outliers available for learning. In this paper, we propose a clustering-based outlier detection method by constructing a model representing a normal data region using only normal data and performing binary classification of outliers and normal data for new data samples. Then, by dividing the given normal data into chunks, and constructing a clustering model for each chunk, we expand it to the ensemble method combining the decision by the models and apply it to the streaming data with dynamic changes. Experimental results using real data and artificial data show high performance of the proposed method.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.1
• /
• pp.434-439
• /
• 2019

Recently, interest in urban temperature change and ground surface temperature change has been increasing due to weather phenomenon due to global warming, heat island phenomenon caused by urbanization in urban areas. In Korea, weather data such as temperature and precipitation have been collected since 1904. In recent years, there are 96 ASOS stations and 494 AWS weather observation stations. However, in the case of terrestrial networks, terrestrial meteorological data except measurement points are predicted through interpolation because they provide point data for each installation point. In this study, to improve the resolution of ground surface temperature measurement, the surface temperature using satellite image was calculated and its applicability was analyzed. For this purpose, the satellite images of Landsat 8 OLI TIRS were obtained for Seoul Metropolitan City by seasons and transformed to surface temperature by applying NASA equation to the thermal bands. The ground measurement data was based on the temperature data measured by AWS. Since the AWS temperature data is station based point data, interpolation is performed by Kriging interpolation method for comparison with Landsat image. As a result of comparing the satellite image base surface temperature with the AWS temperature data, the temperature difference according to the season was calculated as fall, winter, summer, based on the RMSE value, Spring, in order of applicability of Landsat satellite image. The use of that attribute and AWS support starts at $2.11^{\circ}C$ and RMSE ${\pm}3.84^{\circ}C$, which reflects information from the extended NASA.

• Journal of Broadcast Engineering
• /
• v.23 no.6
• /
• pp.855-865
• /
• 2018

Sound event detection is one of the research areas to model human auditory cognitive characteristics by recognizing events in an environment with multiple acoustic events and determining the onset and offset time for each event. DCASE, a research group on acoustic scene classification and sound event detection, is proceeding challenges to encourage participation of researchers and to activate sound event detection research. However, the size of the dataset provided by the DCASE Challenge is relatively small compared to ImageNet, which is a representative dataset for visual object recognition, and there are not many open sources for the acoustic dataset. In this study, the sound events that can occur in indoor and outdoor are collected on a larger scale and annotated for dataset construction. Furthermore, to improve the performance of the sound event detection task, we developed a dual CNN structured sound event detection system by adding a supplementary neural network to a convolutional neural network to determine the presence of sound events. Finally, we conducted a comparative experiment with both baseline systems of the DCASE 2016 and 2017.

• Journal of the Korean GEO-environmental Society
• /
• v.19 no.12
• /
• pp.59-64
• /
• 2018

The damages to the reinforced earth retaining wall are divided into the front wall, foundation, drainage and upper slope. Damage of reinforced earth retaining wall is mainly caused by damage caused by drainage problem in the field. Recently, damage caused by snow removal materials have been occurred. Recently, the amount of snow removal materials used in winter is increasing due to abnormal weather. This chlorides degrades the concrete structure, where the reinforced earth retaining wall was no exception. There has recently been a case in which the front wall of the reinforced earth retaining wall deteriorates due to the chlorides introduced into the back filling portion through the drainage passage. Therefore, in this study, the cause of damages of reinforced earth retaining wall constructed in bridge abutment was analyzed, and an analytical study was conducted on the countermeasure. As a result, it was found that chlorides, which was introduced through the drainage system in the expansion joint of the bridge shift part or the upper structure, is infiltrated into the back part of the reinforced earth retaining wall and damaged. Therefore, it is suggested to improve the drainage system and restored the stiffness of the front wall.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.38 no.6
• /
• pp.851-857
• /
• 2018

In this research, LCA analysis of the manufacturing process of pot bearing for fixed, movable in all directions, movable in one direction was carried out to analyze the environmental load using the LCA methodology. Especially, the water footprint that has been and issue in recent years was analyzed. As a result of LCA, it was analyzed that the contribution of the plate was more than 64.2% in all of the six impact categories in the case of fixed pot bearing base, and more than 94% in the category of resource depletion and photochemical oxidant creation. In the case of all direction pot bearing and one direction pot bearing, the contribution of PTFE was the highest in the global warming and stratospheric ozone depletion, and the contribution by the plate was higher in the other impact categories. The water footprint of each type of pot bearing was analyzed as $22.4m^3\;H_2O\;eq/kg$ for one direction pot bearing, $17.1m^3\;H_2O\;eq/kg$ for fixed pot bearing, and $14.1m^3\;H_2O\;eq/kg$ for all direction pot bearing. As a result of life cycle analysis, the contribution of water use in manufacturing was more than 65% in all three types. The results of this study can be used as basic data for decision making in construction method and material selection of bridges in the future.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.17 no.6
• /
• pp.14-24
• /
• 2018

Pedestrian movement of passengers using smart card within stations can be divided into three types of activities - straight ride and alight, line transfer, and station transfer. Straight ride and alight is transfer activity for which the card terminal and embarking line are identical. In this case, straight ride occurs at the origin station and straight alight occurs at the destination station. Line transfer refers to activity in which the subway line embarked on by the passenger is different from that which is disembarked. Succinctly, line transfer is transfer at a middle station, rather than at origin or destination stations. Station transfer occurs when the card terminal line and embarking line are different. It appears when station transfer happens at the origin station as starting transfer, and at the destination station as destination transfer. In the case of Metropolitan smart card data, origin and destination station card terminal line number data is recorded, but subway line data does not exist. Consequently, transportation card data, as it exists, cannot adequately be used to analyze pedestrian movement as a whole in subway stations. This research uses the smart card data, with its constraints, to propose an analysis model for passenger pedestrian movement within subway stations. To achieve this, a path selection model is constructed, which links origin and destination stations, and then applied for analysis. Finally, a case study of the metropolitan subway is undertaken and pedestrian volume analyzed.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.36 no.6
• /
• pp.601-609
• /
• 2018

The Smart City Plan specifies that it should reflect the content of the Smart City Comprehensive Plan, which is the upper plan while presenting the basic direction, promotion strategy, service establishment and operation plans of a smart city construction project. However, there are not enough empirical studies on whether plan contents are systematically established according to smart city planning hierarchy. In this study, we are to use the social network analysis to find out the local government's smart city plan is actually highly connected to the Smart City Comprehensive Plan, a master plan at national level. To this end, we conducted a social network analysis on Gwangmyeong and Chuncheon, which were recently approved for smart city planning. First, 108 keywords were derived from the 2nd Smart City Comprehensive Plan, and the connectivity between these keywords and Smart City Planning was analyzed. The results of the social network analysis showed that the total connections of Smart City Planning in Gwangmyeong was 371, which was higher than 307 in Chuncheon, and the average degree of connection per keyword and network density were also higher in the Gwangmyeong Smart City Planning than the Chuncheon Smart City Planning. The results of the study showed that the Smart City Planning actually had a high connectivity with the Smart City Comprehensive Plan, and the keywords with high connection centrality were different for each local government. The result of this study can be used as a basis for judging whether there is a high correlation between plans.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.23 no.2
• /
• pp.67-74
• /
• 2019

In this paper the response factor is investigated for middle and small size-RC slab aged bridges. The response factor consists of static and dynamic response factors and is a main parameter in the frequency based-bridge load carrying capacity prediction model. Static and dynamic response factors are determined based on the frequency variation and the impact factor variation respectively between current and previous (or design) states of bridges. Here, the impact factor variation is figured out using the impact factor response spectrum which provides the impact factor according to the natural frequency of bridges. In this study, four actual RC slab bridges aged over 30 years after construction are considered and their span length is 12m. The dynamic loading test in field using a dump truck and eigenvalue analysis with FE models are conducted to identify the current and previous (or design) state-natural frequencies of the bridges, respectively. For more realistic considerations in the moving loading situation, the impact factor response spectrum is developed based on tri-axle moving loads representing the dump truck load distribution and various supporting conditions such as simply supported and both ends fixed conditions. From the results, the response factor is widely ranged from 0.21to 0.91, showing that the static response factor contributes significantly on the results while the dynamic response factor has a small effect on the result. Compared to the results obtained from the impact factor response spectrum based on the single axle-simply supported condition, the maximum percentage difference of the response factors is below 3.2% only.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.23 no.2
• /
• pp.116-121
• /
• 2019

In order to evaluating applicability of RAP (recycled aggregate powder) in mortar, in this study, physical and mechanical tests was carried out. Material characteristics of recycled aggregate and RAP were evaluated and the mechanical properties of mortar replaced with RAP were analyzed. Test result of sieve analysis showed that as the milling time increased the fineness modulus was decreased and the distribution of 0.6 mm particle size was found to increase. The fluidity of mortar mixture substituted with RAP tended to increase than Plain mixture. It was result that the increasing fluidity was affected by unreacted surplus water in the mortar as the binder was replaced with RAP. From the compressive strength result of the mortar subjected to RAP, it was found that the RAP was able to replace up to about 10% of unit binder weight although the compressive strength of mortar was decreased as the RAP replacement increased. From the above study, it can be concluded that the physical properties of RAP satisfied the quality standard of aggregate for replacement with fine aggregate. Moreover, in case of the RAP was replaced up to 10% of unit cement weight, it was able to be possible to improve fluidity and compressive strength of mortar.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.23 no.5
• /
• pp.1-12
• /
• 2019

The purpose of this study is to enhance the seismic performance of exposed column-base plate weak-axis connections for small-sized steel buildings. According to the site inspection for the small-sized building construction, the arbitrary connection details in steel buildings have been applied at the job site, which is considered to be insufficient to guarantee structural safety and stability considering the increased seismic risk in Korea. Therefore, a series of test programs had been done to develop enhanced connection details in order to ensue the adequate seismic stability and safety of small-sized steel buildings. From the test results, It was found that the exposed column-base plate weak-axis connections commonly used in Korea shows very poor seismic behavior due to the "Rocking" phenomena caused by the residual plastic deformation of anchor bolts between anchor plate and concrete. A series of hysteretic tests for finding that solution were conducted to reduce the "Rocking" phenomena of the column-base plate connections, and local buckling of webs in H-column. Finally the enhanced stable seismic behavior was obtained by reinforcing at least 8 anchor bolts with good bonding strength and stiffeners to the webs in H-column.