• Smart Structures and Systems
• /
• v.33 no.4
• /
• pp.291-300
• /
• 2024

Uncertainty of the model, system delay and drive dynamics can be considered as normal uncertainties, and the main source of uncertainty in the seismic control system is related to the nature of the simulated seismic error. In this case, optimizing the management strategy for one particular seismic record will not yield the best results for another. In this article, we propose a framework for online management of active structural management systems with seismic uncertainty. For this purpose, the concept of reinforcement learning is used for online optimization of active crowd management software. The controller consists of a differential controller, an unplanned gain ratio, the gain of which is enhanced using an online reinforcement learning algorithm. In addition, the proposed controller includes a dynamic status forecaster to solve the delay problem. To evaluate the performance of the proposed controllers, thousands of ground motion data sets were processed and grouped according to their spectrum using fuzzy clustering techniques with spatial hazard estimation. Finally, the controller is implemented in a laboratory scale configuration and its operation is simulated on a vibration table using cluster location and some actual seismic data. The test results show that the proposed controller effectively withstands strong seismic interference with delay. The goals of this paper are towards access to adequate, safe and affordable housing and basic services, promotion of inclusive and sustainable urbanization and participation, implementation of sustainable and disaster-resilient buildings, sustainable human settlement planning and manage. Simulation results is believed to achieved in the near future by the ongoing development of AI and control theory.

• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.1-21
• /
• 2020

With the rapid development of artificial intelligence technology, various techniques have been developed to extract meaningful information from unstructured text data which constitutes a large portion of big data. Over the past decades, text mining technologies have been utilized in various industries for practical applications. In the field of business intelligence, it has been employed to discover new market and/or technology opportunities and support rational decision making of business participants. The market information such as market size, market growth rate, and market share is essential for setting companies' business strategies. There has been a continuous demand in various fields for specific product level-market information. However, the information has been generally provided at industry level or broad categories based on classification standards, making it difficult to obtain specific and proper information. In this regard, we propose a new methodology that can estimate the market sizes of product groups at more detailed levels than that of previously offered. We applied Word2Vec algorithm, a neural network based semantic word embedding model, to enable automatic market size estimation from individual companies' product information in a bottom-up manner. The overall process is as follows: First, the data related to product information is collected, refined, and restructured into suitable form for applying Word2Vec model. Next, the preprocessed data is embedded into vector space by Word2Vec and then the product groups are derived by extracting similar products names based on cosine similarity calculation. Finally, the sales data on the extracted products is summated to estimate the market size of the product groups. As an experimental data, text data of product names from Statistics Korea's microdata (345,103 cases) were mapped in multidimensional vector space by Word2Vec training. We performed parameters optimization for training and then applied vector dimension of 300 and window size of 15 as optimized parameters for further experiments. We employed index words of Korean Standard Industry Classification (KSIC) as a product name dataset to more efficiently cluster product groups. The product names which are similar to KSIC indexes were extracted based on cosine similarity. The market size of extracted products as one product category was calculated from individual companies' sales data. The market sizes of 11,654 specific product lines were automatically estimated by the proposed model. For the performance verification, the results were compared with actual market size of some items. The Pearson's correlation coefficient was 0.513. Our approach has several advantages differing from the previous studies. First, text mining and machine learning techniques were applied for the first time on market size estimation, overcoming the limitations of traditional sampling based- or multiple assumption required-methods. In addition, the level of market category can be easily and efficiently adjusted according to the purpose of information use by changing cosine similarity threshold. Furthermore, it has a high potential of practical applications since it can resolve unmet needs for detailed market size information in public and private sectors. Specifically, it can be utilized in technology evaluation and technology commercialization support program conducted by governmental institutions, as well as business strategies consulting and market analysis report publishing by private firms. The limitation of our study is that the presented model needs to be improved in terms of accuracy and reliability. The semantic-based word embedding module can be advanced by giving a proper order in the preprocessed dataset or by combining another algorithm such as Jaccard similarity with Word2Vec. Also, the methods of product group clustering can be changed to other types of unsupervised machine learning algorithm. Our group is currently working on subsequent studies and we expect that it can further improve the performance of the conceptually proposed basic model in this study.

• Journal of Korean Society for Atmospheric Environment
• /
• v.28 no.3
• /
• pp.233-248
• /
• 2012

The national emission from energy sector accounted for 84.7% of all domestic emissions in 2007. Of the energy-use emissions, the emission from mobile source as one of key categories accounted for 19.4% and further the road transport emission occupied the most dominant portion in the category. The road transport emissions can be estimated on the basis of either the fuel consumed (Tier 1) or the distance travelled by the vehicle types and road types (higher Tiers). The latter approach must be suitable for simultaneously estimating $CO_2$, $CH_4$, and $N_2O$ emissions in local administrative districts. The objective of this study was to estimate 31 municipal GHG emissions from road transportation in Gyeonggi Province, Korea. In 2008, the municipalities were consisted of 2,014 towns expressed as Dong and Ri, the smallest administrative district unit. Since mobile sources are moving across other city and province borders, the emission estimated by fuel sold is in fact impossible to ensure consistency between neighbouring cities and provinces. On the other hand, the emission estimated by distance travelled is also impossible to acquire key activity data such as traffic volume, vehicle type and model, and road type in small towns. To solve the problem, we applied a hierarchical cluster analysis to separate town-by-town road patterns (clusters) based on a priori activity information including traffic volume, population, area, and branch road length obtained from small 151 towns. After identifying 10 road patterns, a rule building expert system was developed by visual basic application (VBA) to assort various unknown road patterns into one of 10 known patterns. The expert system was self-verified with original reference information and then objects in each homogeneous pattern were used to regress traffic volume based on the variables of population, area, and branch road length. The program was then applied to assign all the unknown towns into a known pattern and to automatically estimate traffic volumes by regression equations for each town. Further VKT (vehicle kilometer travelled) for each vehicle type in each town was calculated to be mapped by GIS (geological information system) and road transport emission on the corresponding road section was estimated by multiplying emission factors for each vehicle type. Finally all emissions from local branch roads in Gyeonggi Province could be estimated by summing up emissions from 1,902 towns where road information was registered. As a result of the study, the GHG average emission rate by the branch road transport was 6,101 kilotons of $CO_2$ equivalent per year (kt-$CO_2$ Eq/yr) and the total emissions from both main and branch roads was 24,152 kt-$CO_2$ Eq/yr in Gyeonggi Province. The ratio of branch roads emission to the total was 0.28 in 2008.

• Journal of Korean Society of Forest Science
• /
• v.80 no.4
• /
• pp.393-407
• /
• 1991

The vegetation structure within Chamaecyparis obtusa plantation was analyzed for the purpose of applying the effective forestation method for Chmaecyparis obtusa plantation, tending and regeneration in the southern districts of korea. The results were as follows ; 1. The importance percentage was high in the order of Eurya japonica, Rhus verniciflua, Chamaecyparis obtusa, Lindera erythrocarpa, Carpinus laxiflora, Styrax japonica, Viburnum dilatatum, Zanthoxylum piperitum and Smilax china among the vegetation of Chamaecyparis obtusa. Importance percentage of natural seedling of Chamaecyparis obtusa was high in lower story but gradually decreased in middle story. 2. The basal area of upper trees had a negative correlation with the density of natural seedlings in the middle and lower story, and it represents that the basal area of upper trees had some effect on the density of natural seedlings within understories. 3. The rate of the A and B class by Raunkiaer's frequency was higher in the vegetation of middle story than that of lower story. 4. By Morisita's index, the species of Chamaecyparis obtusa, Rhus verniciflua, Lindera erythrocarpa, Smilax china. Callicarpa japonica and Lindera obtusiloba were randomly distributed at lower story, but they were aggregatively distributed at middle story. At all of middle and lower story, Eurya japonica and Viburum dilatatum were randomly distributed, and Carpinus laxiflora, Zanthoxylum piperitum and Picrasma quassioides were aggregatively distributed. 5. The number of appearance species and the value of species diversity in western survey area were more than that of eastern survey area. 6. The value of species diversity at lower story was higher than that of middle story, and it represents that the number of individuals of appearance species was composed more even at lover story than middle story. 7. According to cluster analysis by similarity index, the survey areas were separated from inland and seacoast districts. 8. Judging from each stories ordination analysis by dissimilarity index, the vegetation was separated from lower and middle story, and the vegetation of lower story was more progressed succession stage than that of middle story. 9. In Chamaecyparis obtusa stands, Eurya japonica had a positive correlation with Sorbus alnifolia, Hex macropoda. Ficus erecta and Trachelospermum asiaticum, but it had a negative correlation with Zanthoxylum piperitum, Carpinus laxiflora and Parthenocissus tricuspidata. 10. In estimation of the productivity of Chamaecyparis obtusa stands, the value of SC (Conic surface) and VP (Parabolic volume) for upper trees was 94.5% and 99.63%, respectively and SC and VP of middle story was 5.49% and 0.37%, respectively. In the species of middle story, material productivity was high in order of Eurya japonica. Lindera eryhrocarpa, Rhus verniciflua. Carpinus laxiflora and Styrax japonica.

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.49 no.6
• /
• pp.516-521
• /
• 2004

The straw of thirty one rice varieties were evaluated for their feed value and related agronomic traits. The rice straw were hand-harvested, dried to constant weight at $75^{\circ}C$ and ground through a 20 mesh seive in a Wiley mill, analyzed with crude protein (CP), acid detergent fiber (ADF) and neutral detergent fiber (NDF). Relative feed value (RFV) was calculated from NDF and ADF. The sum of standardized score was estimated by dry weight of rice straw, content of CP, ADF and NDF. The straw yield of Daeanbyeo was 725.9 kg/10a, showed heighest value among the varieties and remainder was in the order of Keumnambyeo, Donginbyeo #1 and Chucheongbyeo. Crude protein (CP) content in a Dasanbyeo was higher than those in other varieties. The content of ADF in a Junghwabyeo and NDF in a Sobaegbyeo were $34.3\%$ and $63.8\%$, respectively, showed lowest value among the varieties. The rice straw of Dunnaebyeo, Obongbyeo, Seoanbyeo, Keumobyeo, Hwaseongbyeo, Noganbyeo and Gyehwabyeo belonged to the high feed value varieties by estimation of cluster analysis, sum of standardized score and RFV. The content of CP was found to be positively related with dry weight of leaf and grain, but negatively related with heading days after seeding, culm length, specific leaf weight (SLW) and dry weight of stem. ADF and NDF were found to be positively related with heading days after seeding, culm length, SLW and dry weight of leaf, but negatively related with dry weight of stem. The sum of standardized score and RFV were the only positive relationship with dry weight of stem and negative relationship with other traits.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.23 no.1
• /
• pp.50-56
• /
• 2011

To make conservative port and coast ecosystems and creative the greenspace, We were investigated with characteristic of flora, environmental factors, types of port, adaptive species, minimum conservation area and plantation model. In 50 sites of study areas, there are 19 families and 174 species of vascular plants and 19 families and 48 species of halophytes. Dominant communities in port ecosystem contains Carex kobomugi community, Elymus mollis community, Carex pumila community, Ixeris repens community, Vitex rutundifolia community, Calystegia soldandlla community, Rosa rugosa community, Lathyrus japonica community, Salsola komarovi community, Cynodon dactylon community, Tetragonia tetragonioides community, Suaeda japonica community, Suaeda maritima community, Zoysia sinica community and Phragmites communis community. We carried out Canonical Correspondence Analysis(CCA) for ordinations on the vegetation and plant communities-environmental variable matrices in 50 sites. The communities tended to cluster into three types: Clay marsh, Sand marsh, Sand gravel marsh types. Adaptive species in habitate types are selected that sand marsh-type communities in ports contained Elymus mollis community, Ixeris repens community, Carex kobomugi community, Carex pumila community, Clay marsh-type communities contained Suaeda japonica community, Phragmites communis community, Zoysia sinica community and Suaeda maritima community, Sand gravel marsh-type communities contained Vitex rutundifolia community, Calystegia soldandlla community. We are conducted the estimation of minimal area for plantation of adaptive plant species and carried out guide line and plantation model for creation of green space in port ecosystem.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.13 no.1
• /
• pp.177-185
• /
• 2009

Cluster analysis or clustering is a kind of unsupervised learning method in which a set of data points is divided into a given number of homogeneous groups. Fuzzy clustering method, one of the most popular clustering method, allows a point to belong to all the clusters with different degrees, so produces more intuitive and natural clusters than hard clustering method does. Even more some of fuzzy clustering variants have noise-immunity. In this paper, we improved the Possibilistic Fuzzy C-Means (PFCM), which generates a membership matrix as well as a typicality matrix, using Gath-Geva (GG) method. The proposed method has a focus on the boundaries of clusters, which is different from most of the other methods having a focus on the centers of clusters. The generated membership values are suitable for the classification-type applications. As the typicality values generated from the algorithm have a similar distribution with the values of density function of Gaussian distribution, it is useful for Gaussian-type density estimation. Even more GG method can handle the clusters having different numbers of data points, which the other well-known method by Gustafson and Kessel can not. All of these points are obvious in the experimental results.

• Journal of Korean Society of Forest Science
• /
• v.103 no.2
• /
• pp.196-202
• /
• 2014

In order to select superior trees based on fruit characteristics and provide basic informations necessary for their improvement, total 107 individual trees of Sorbus alnifolia (Sieb. et Zucc.) K. Koch were selected from 11 wild populations in South Korea. After collecting normal fruit branch, we investigated morphological characteristics of fruit and then considered its relationship among the 11 populations by multivariate analysis method. Results from principal compound analysis showed that it represented 85.8% accumulated explanation from five principal compounds. According to cluster analysis based on fruit characteristics, the natural S. alnifolia populations were classified into four groups and Mt. Mani population was different from other populations. Selection effect with outstanding candidate trees including superior 5 individual trees (Gwangyo 1, Gwangyo 2, Deogyu 7, Mani 29, Mani 30) was estimated at 122.8%, 115.5% and 182.7% in fruit width, length and yield per fruit bunch, respectively. The object of this results will give us invaluable information about breeding by selection of S. alnifolia in south Korea.

• Journal of Korean Society of Transportation
• /
• v.36 no.3
• /
• pp.203-215
• /
• 2018

This study suggests a method to construct large scale dynamic O/D reflecting the characteristic that the passengers' travel patterns change according to the land use patterns of the destination. There are limitations in the existing research about dynamic O/D estimation method, such as the difficulty of collecting data, which can be applied only to a small area, or limiting to a specific transportation network such as highway networks or public transportation networks. In this paper, we propose a method to estimate dynamic O/D without limitation of analysis area based on transportation resources that can be easily collected and used according to the big data era. Clustering analysis was used to calculate the departure time trip distribution ratio based on arrival time and departure time trip distribution function was estimated by each cluster. As a result of the comparison test with the survey data, the estimated distribution function was statistically significant.

• The Journal of the Acoustical Society of Korea
• /
• v.30 no.8
• /
• pp.421-427
• /
• 2011

This study aimed to figure out the numbers and note distributions and sexagenary cycles of old pyeongyoungs systematically, and estimate the nominal frequency of whangjongeum, the Korean tradition pitch standard. As a total 214 old stones in the National Palace Museum, the National Kukak Center, the Kukak National High School were counted by notes and sexagenaries. The nominal frequencies of 17 old whangjong stones' sounds were categorized by cluster analysis method. Using nominal frequencies of stones according to their sexagenaries and Korean traditional intonation were used to estimate the nominal frequencies of the whangjong. The nominal frequency can be estimated by 22 Keychuk stones as 266.9 Hz, by Cheongyu and Gabja stones as 262.4~262.5 Hz, and by Gabjin, Sowha 12 and Sowha 13 as Estimating by 22 Kyechuk stones which were matched with the records. These results seem to be more reliable, because it is based on the whol samples of old pyeongyoungs, while the former studies have been based on couples of whangjong stones' sounds.