• Journal of Korea Society of Industrial Information Systems
• /
• v.28 no.6
• /
• pp.11-20
• /
• 2023

The cost of medical treatment for motor vehicle accidents is increasing every year. In this study, we created a model to predict long-term hospitalization(more than 18 days) among minor patients, which is the main item of increasing traffic accident medical expenses, using five algorithms such as decision tree, and analyzed the factors affecting long-term hospitalization. As a result, the accuracy of the prediction models ranged from 91.377 to 91.451, and there was no significant difference between each model, but the random forest and XGBoost models had the highest accuracy of 91.451. There were significant differences between models in the importance of explanatory variables, such as hospital location, name of disease, and type of hospital, between the long-stay and non-long-stay groups. Model validation was tested by comparing the average accuracy of each model cross-validated(10 times) on the training data with the accuracy of the validation data. To test of the explanatory variables, the chi-square test was used for categorical variables.

• Journal of Korean Society of Forest Science
• /
• v.102 no.1
• /
• pp.38-58
• /
• 2013

The existence of endemism in many parts of the world is an important factor for conservationists. Conservation can only be carried out under national legislation, and national endemics, which have very limited ranges, fully depend on the effort and success of conservation. A total of 523 vascular plant taxa were listed in the latest national checklist by Ministry of Environment in 2005, while the 'Creation and Furtherance of Arboretums Act' including a national endemic list (appendix 4-1) was established by Korea Forest Service and was legislated as a law in late 2011. This legislation by Korea Forest Service on endemism of Korean vascular plants have required much attention because of discrepancies of nomenclature, taxonomic bias and inflation. Examining data for both lists proposed by Ministry of Environment and Korea Forest Service, of the total of 360 legislated taxa, around 286 taxa are shared with the list of Ministry of Environment, of which about 80% have been found as common taxa. Around 67(18.7%) are typographic errors, and 14 taxa (3.9%) are recorded as illegitimate and invalidly published names. Through this analysis 12 taxa (3.4%) were found in China as well as in Korea and these are thought to be non Korean endemic taxa. Taken together, the legislated list displayed 1/4 (24.9%) errors out of the total list. Only 59 taxa (16.5%) are identified as national endemic species. The remainder are either unresolved candidates (73 taxa, 20.4%) or synonyms (196 taxa, 54.7%) status. It must be noted, that the concept of endemism very much depends on the knowledge of the species concept, taxonomic bias and geographical range of a species. Also, the most major nomenclatural problem tend to be more stable if the information on database about Korea Plant Name, which is managed by Korea National Arboretum are well updated year to year. These exaggerated numbers underscore the urgency for regional conservation planning and implementing effective strategies to preserve these real endemic taxa into the future.

• Korean Journal of Plant Taxonomy
• /
• v.34 no.1
• /
• pp.37-41
• /
• 2004

T. Nakai, who wrote Flora of Koreana vol. II in 1911, cited V. L. Komarov's collections and listed more than 120 specimens with ca. 65 localities. All collection sites cited were described in the Romanized characters based on the Russian pronunciation. Therefore, it is very difficult to pinpoint those sites using the current or old version of Korean map. From this study, many names were reviewed based on other studies (both the Russian version and the Japanese translated version) and records (e.g. specimen label) and presented based on the current provinces with local names, the Chinese character, the GPS data and the first page of citation.

• The Korean Journal of Quaternary Research
• /
• v.17 no.2
• /
• pp.75-78
• /
• 2003

This study is conducted to examine the solar radiation features of landscape in Eastern Mongolia and their space and time distribution characteristics. To many foreigners, the name Mongolia conjures up images of the vast steppes of Central Asia. And, indeed, the extensive grasslands of the steppe make up the heart of Mongolia-geographically and economically. Steppe covers nearly the entire far eastern part of Mongolia, extending west in a narrowing band just south of the Khangai and Khan Khokhii mountains all the way to the Depression of the Great lakes. Eastern Mongolia is recognized as one of the last remaining untouched grassland steppes in the world. Landscape of Eastern Mongolia divided by 19 types that belongs to High Mountain, Middle Mountain and low mountain steppe. Only 5.6% of the total land area is covered by forest. Eastern Mongolian steppe is homeland of migratory rare and endangered birds and about 2.0million freely migrating gazelles. According to Mongolian scientists, there is growing evidence of fundamental changes in the Eastern steppe ecosystems in terms of increasing aridity during the last 70 years. As estimated average annual temperature has increased by approximately 0.7oC, soil moisture, energy supply has been changing. These processes closely interrelated into water and energy cycle of steppe ecosystem.

• Mycobiology
• /
• v.46 no.3
• /
• pp.192-204
• /
• 2018

The name Golovinomyces cynoglossi s. lat. is traditionally applied to a complex of morphologically similar powdery mildews on hosts of the plant family Boraginaceae. The current species-level taxonomy within this complex is ambiguous due to the lack of phylogenetic examinations. The present study applied phylogenetic methods to clarify the taxonomy of G. cynoglossi s. lat. Phylogenetic analysis of rDNA ITS sequences retrieved from Asian, European and North American specimens revealed that G. cynoglossi s. lat. collections from different hosts involved several species in five clearly separated lineages. Clade I consists primarily of Golovinomyces cynoglossi s. str. on Cynoglossum. Clade III consists of Golovinomyces sequences retrieved from the host genera Symphytum and Pulmonaria. The taxa within clade III are now assigned to G. asperifoliorum comb. nov. Clade V encompasses G. cynoglossi s. lat. on the host genera Bothriospermum, Buglossoides, Echium, Myosotis, and Trigonotis. The taxa within clade V are now assigned to G. asperifolii comb. nov. The species concerned in this study were lecto- and epitypified to stabilize their nomenclature.

• Korean Journal of Plant Taxonomy
• /
• v.41 no.4
• /
• pp.361-364
• /
• 2011

We report an unrecorded plant species, Elatostema laetevirens Makino (Urticaceae), in Korea. This species was collected for the first time from under the canopy of an evergreen forest in Hyodoncheon, Sanghyo-dong, Soeguipo-si and Tamra valley in Ohdeung-dong, Cheju-si of the Province of Cheju in Korea. E. laetevirens Makino was quite distinct from others of the genus Elatostema in Korea owing to its perennial nature, the absence of trichomes in the stem and the absence of peduncles during the inflorescence period. This species was given the Korean name 'pu-reun-mong-ul-pul' based on the absence of luster and the bluish green color on the adaxial surface of its leaves.

• Korean Journal of Plant Taxonomy
• /
• v.40 no.4
• /
• pp.262-266
• /
• 2010

A new variety of Elsholtzia splendens (Lamiaceae) was found by the roadside of a mountain base at Jungmun-dong, Seogwipo-si, Jeju-do. The new taxon is termed Elsholtzia splendens var. fasciflora N.S. Lee, M.S. Chung & C.S. Lee, var. nov. The Korean name 'Da-bal-kkot-hyang-yu' was given. The morphological characters and illustrations of the species based on the holotype are provided together with a photograph of the habitat. It is distinctly distinguished from Elsholtzia splendens var. splendens by a few characteristics: leaves that area 2-4 cm long, leaves beneath the inflorescence, typically mostly 4; and inflorescence mosyly 2-4, fascicled, involucres oval.

• Korean Journal of Plant Taxonomy
• /
• v.38 no.4
• /
• pp.557-563
• /
• 2008

일본, 중국, 대만 및 베트남에 분포하는 것으로 알려진 부처손과의 각시비늘이끼 (Selaginella heterostachys Baker)가 강원도 춘천시 남산면 강촌리 저지대에서 발견되었다. 이 종은 선비늘이끼와 유사하나 측면의 소엽이 난상이며 기부가 심장저이고, 중앙부의 소엽은 길며 정단부가 구부러진 점으로 구별된다. 국명은 선비늘이끼보다 중앙부의 소엽이 구부러져 각시의 치마와 비슷하여 각시비늘이끼라고 신칭하였다. 각시비늘이끼의 형태적 특징을 기재하였고, 도해와 생태 사진을 첨부하였다.

• Korean Journal of Plant Taxonomy
• /
• v.45 no.2
• /
• pp.136-144
• /
• 2015

We reviewed the taxonomic identity and morphological characteristics of Cardamine manshurica (Kom.) Nakai based on its original description and type specimens. Cardamine manshurica, with its smaller flowers and siliques and earlier flowering season, is distinguished from the closely related species C. flexuosa With. This species-level difference is additionally supported in a phylogenetic tree constructed using nrDNA ITS sequences. Taken together, we therefore propose that C. manshurica is a distinct species. The scientific name 'Cardamine manshurica (Kom.) Nakai' is accepted.

• Journal of Korea Multimedia Society
• /
• v.19 no.9
• /
• pp.1647-1658
• /
• 2016

RGB+D database has been widely used in object recognition, object tracking, robot control, to name a few. While rapid advance of active depth sensing technologies allows for the widespread of indoor RGB+D databases, there are only few outdoor RGB+D databases largely due to an inherent limitation of active depth cameras. In this paper, we propose a novel method used to build outdoor RGB+D databases. Instead of using active depth cameras such as Kinect or LIDAR, we acquire a pair of stereo image using high-resolution stereo camera and then obtain a depth map by applying stereo matching algorithm. To deal with estimation errors that inevitably exist in the depth map obtained from stereo matching methods, we develop an approach that estimates confidence of depth maps based on unsupervised learning. Unlike existing confidence estimation approaches, we explicitly consider a spatial correlation that may exist in the confidence map. Specifically, we focus on refining confidence feature with the assumption that the confidence feature and resultant confidence map are smoothly-varying in spatial domain and are highly correlated to each other. Experimental result shows that the proposed method outperforms existing confidence measure based approaches in various benchmark dataset.