• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.135-149
• /
• 2020

Artificial intelligences are changing world. Financial market is also not an exception. Robo-Advisor is actively being developed, making up the weakness of traditional asset allocation methods and replacing the parts that are difficult for the traditional methods. It makes automated investment decisions with artificial intelligence algorithms and is used with various asset allocation models such as mean-variance model, Black-Litterman model and risk parity model. Risk parity model is a typical risk-based asset allocation model which is focused on the volatility of assets. It avoids investment risk structurally. So it has stability in the management of large size fund and it has been widely used in financial field. XGBoost model is a parallel tree-boosting method. It is an optimized gradient boosting model designed to be highly efficient and flexible. It not only makes billions of examples in limited memory environments but is also very fast to learn compared to traditional boosting methods. It is frequently used in various fields of data analysis and has a lot of advantages. So in this study, we propose a new asset allocation model that combines risk parity model and XGBoost machine learning model. This model uses XGBoost to predict the risk of assets and applies the predictive risk to the process of covariance estimation. There are estimated errors between the estimation period and the actual investment period because the optimized asset allocation model estimates the proportion of investments based on historical data. these estimated errors adversely affect the optimized portfolio performance. This study aims to improve the stability and portfolio performance of the model by predicting the volatility of the next investment period and reducing estimated errors of optimized asset allocation model. As a result, it narrows the gap between theory and practice and proposes a more advanced asset allocation model. In this study, we used the Korean stock market price data for a total of 17 years from 2003 to 2019 for the empirical test of the suggested model. The data sets are specifically composed of energy, finance, IT, industrial, material, telecommunication, utility, consumer, health care and staple sectors. We accumulated the value of prediction using moving-window method by 1,000 in-sample and 20 out-of-sample, so we produced a total of 154 rebalancing back-testing results. We analyzed portfolio performance in terms of cumulative rate of return and got a lot of sample data because of long period results. Comparing with traditional risk parity model, this experiment recorded improvements in both cumulative yield and reduction of estimated errors. The total cumulative return is 45.748%, about 5% higher than that of risk parity model and also the estimated errors are reduced in 9 out of 10 industry sectors. The reduction of estimated errors increases stability of the model and makes it easy to apply in practical investment. The results of the experiment showed improvement of portfolio performance by reducing the estimated errors of the optimized asset allocation model. Many financial models and asset allocation models are limited in practical investment because of the most fundamental question of whether the past characteristics of assets will continue into the future in the changing financial market. However, this study not only takes advantage of traditional asset allocation models, but also supplements the limitations of traditional methods and increases stability by predicting the risks of assets with the latest algorithm. There are various studies on parametric estimation methods to reduce the estimated errors in the portfolio optimization. We also suggested a new method to reduce estimated errors in optimized asset allocation model using machine learning. So this study is meaningful in that it proposes an advanced artificial intelligence asset allocation model for the fast-developing financial markets.

• Journal of Intelligence and Information Systems
• /
• v.26 no.3
• /
• pp.149-169
• /
• 2020

"The Urban Renewal New Deal project", one of the government's major national projects, is about developing underdeveloped areas by investing 50 trillion won in 100 locations on the first year and 500 over the next four years. This project is drawing keen attention from the media and local governments. However, the project model which fails to reflect the original characteristics of the area as it divides project area into five categories: "Our Neighborhood Restoration, Housing Maintenance Support Type, General Neighborhood Type, Central Urban Type, and Economic Base Type," According to keywords for successful urban regeneration in Korea, "resident participation," "regional specialization," "ministerial cooperation" and "public-private cooperation", when local governments propose urban regeneration projects to the government, they can see that it is most important to accurately understand the characteristics of the city and push ahead with the projects in a way that suits the characteristics of the city with the help of local residents and private companies. In addition, considering the gentrification problem, which is one of the side effects of urban regeneration projects, it is important to select and implement urban regeneration types suitable for the characteristics of the area. In order to supplement the limitations of the 'Urban Regeneration New Deal Project' methodology, this study aims to propose a system that recommends urban regeneration types suitable for urban regeneration sites by utilizing various machine learning algorithms, referring to the urban regeneration types of the '2025 Seoul Metropolitan Government Urban Regeneration Strategy Plan' promoted based on regional characteristics. There are four types of urban regeneration in Seoul: "Low-use Low-Level Development, Abandonment, Deteriorated Housing, and Specialization of Historical and Cultural Resources" (Shon and Park, 2017). In order to identify regional characteristics, approximately 100,000 text data were collected for 22 regions where the project was carried out for a total of four types of urban regeneration. Using the collected data, we drew key keywords for each region according to the type of urban regeneration and conducted topic modeling to explore whether there were differences between types. As a result, it was confirmed that a number of topics related to real estate and economy appeared in old residential areas, and in the case of declining and underdeveloped areas, topics reflecting the characteristics of areas where industrial activities were active in the past appeared. In the case of the historical and cultural resource area, since it is an area that contains traces of the past, many keywords related to the government appeared. Therefore, it was possible to confirm political topics and cultural topics resulting from various events. Finally, in the case of low-use and under-developed areas, many topics on real estate and accessibility are emerging, so accessibility is good. It mainly had the characteristics of a region where development is planned or is likely to be developed. Furthermore, a model was implemented that proposes urban regeneration types tailored to regional characteristics for regions other than Seoul. Machine learning technology was used to implement the model, and training data and test data were randomly extracted at an 8:2 ratio and used. In order to compare the performance between various models, the input variables are set in two ways: Count Vector and TF-IDF Vector, and as Classifier, there are 5 types of SVM (Support Vector Machine), Decision Tree, Random Forest, Logistic Regression, and Gradient Boosting. By applying it, performance comparison for a total of 10 models was conducted. The model with the highest performance was the Gradient Boosting method using TF-IDF Vector input data, and the accuracy was 97%. Therefore, the recommendation system proposed in this study is expected to recommend urban regeneration types based on the regional characteristics of new business sites in the process of carrying out urban regeneration projects."

• Journal of Korean Society of Forest Science
• /
• v.30 no.1
• /
• pp.30-41
• /
• 1976

The workers involved in breeding the growth of forest trees as well as in the efficiency of breeding work wish to find out the evidence that juvenile-mature correlations are high and that therefore early tests can be profitably employed in relation work. Juvenile-mature correlation denotes in general sense the interdependence between qualitative or quantitative data collected at different intervals during the life cycle. The correlation can also be obtained through the stem analysis if there is the possibility of cutting sample trees needed amount. In the present study, the juvenile-mature correlation coefficients are calculated from the stem analysis data. The every possible values of correlation coefficient between the 5-year age groups as to diameter, height, and volume growth of Pinus koraiensis S. et Z. and Larix leptolepis Gord. grown in the middle district of Korea were calculated. All the sample trees were cut from the man made plantation. The correlation coefficients are presented in tables and figures. In Pinus koraiensis S. et Z., the values of correlation coefficient between the successive age groups of heights growth are lower in general than those values pertaining to diameter growth. This tendencies are indifferent to site quality. In Larix leptolepis Gord., the values were lower thant these of Pinus koraiensis S.etZ. In any species and characteristics studied, the implications that at least 15 years growth character are related to 35 or 40 years size with reliability could be deduced. Through the relative ranking studies of diameter, height and volume growth of Larix leptolepis Gord., the large varieties among individual trees are appeared during 5-year to 35 year old.

• Journal of Korean Society of Forest Science
• /
• v.102 no.4
• /
• pp.578-586
• /
• 2013

This study was performed in Pinus rigida and Larix kaempferi plantations which occupy approximately 60% of artificial forest area in Korea. The objective of this study was to know the differences in soil physical and chemical properties between both plantations. Soil physical and chemical properties from published literature and analyzed soil data in national forest in 2010 and 2011 were analyzed by plantation regions and stand age of 5 years unit. Jeollanamdo in Pinus rigida plantations and Gyeongsangbuk-do in Larix kaempferi plantations showed higher soil chemical properties than those of other regions. Soil texture in both plantations was almost loam and sandy loam. Mean soil pH in Pinus rigida and Larix kaempferi plantations were 4.86 and 4.87, respectively and there was no relationship between soil pH and stand age. The mean concentrations of total nitrogen (%) and available phosphorus (mg $kg^{-1}$) were 0.21 and 11.00 for Pinus rigida plantation and 0.28 and 13.32 for Larix kaempferi plantation. In Larix kaempferi plantation, total nitrogen, available phosphorus and organic matter concentrations and C.E.C. were higher than those in Pinus rigida plantation and showed positive relationship with stand age. This positive relationship was also revealed between the exchangeable cations and soil pH. The results of this study provide an informative data in selecting tree species for planting and contribute to the establishing forest management plan for the maintenance of sustainable forests resources.

• Journal of Korea Multimedia Society
• /
• v.9 no.7
• /
• pp.872-893
• /
• 2006

In this study, I classify the elements of physiognomic judgment of character with regard to form and meaning from a visual perspective based on physiognomic judgment of character study in 'the cosmic dual forces and the Five Elements theory'. Individual characters for each type are designed using graphic data. Based on that, design system of individual characters for each personality type is investigated using Neural Network system. Faces with O-Haeng (Five Elements) shapes are shown to constitute the system with ${\pm}0.3%$ degree of error tolerance for the non-loaming input data. For the shapes of Chinese characters 'tree, fire, soil, gold and water', their MSE(Mean Square Error) are 0.3, 0.3, 0.2, 0.5, 0.2. It seems to be the best regarding the scoring system which ranges from 0 to 5. Therefore, this system might be regarded to produce the most accurate facial shape of character automatically when we input character's personality we desire to make.

• Journal of Plant Biology
• /
• v.24 no.1
• /
• pp.47-57
• /
• 1981

Girth at breast height was measured to test skewness ($g_1$) and kurtosis ($g_2$) of frequency distribution of the basal area in terms of t-test and Kolmogorov-Smirnov test for a total of forty six monocultures within Sudong and Kwhangnung area in central part of Korean peninsula in 1979 and 1980. The monocultures are about 10 to 50 years old, and four kinds: Pinus koraiensis, Larix kaempferi, Abies holophylla and Pinus rigida. Most of the sample sizes per site were ranged 70 to 110 excluding 4 sites. The number of classes interval was based on Sturges rule for each monoculture and was ranged from 5 to 10. In Sudong the range of age(yr) and basal area (($cm^2$)/tree) of the monocultures were from 10 to 20 and from 27.60 to 383. for Kwhangnung they were from 15 to 47 and mostly 102.15 to 619.14, respectively. All 43 monocultures except 1 showed +$g_1$, which ranged from 0.3 to 2.2 except six sites. Of the total 46 sites, 23 sites showed significant +$g_1$ which includes about 10 year-old monoculture. The number of classes interval with significant positive skewness ranged from 6 to 9. The data suggest that intraspecific competition in terms of stand structure seems to appear from about 10 year-old monocultures, and it may even last to about 50 year-old one. Around 24 monocultures showed nonsignificant -$g_2$ except one. Most -$g_2$ ranged from -0.12 to -0.83. Around 20 monocltures showed positive $g_2$ ranging from +0.13 to +3.841. Of the 22 +$g_1$, majority of 11 were very highly significant. Of all monocultures only 5 showed significant result from Kolmogorov-Smirnov test. Of the 4 species, Larix kaempferi seems to show density stress first then Abies holophylla, and Pinus koraiensis last. Data of this study indicate that adequate number of classes intervals and sample sizes for studying intraspecific competition in terms of basal area are 6 to 9 and 80 trees rather than 12 and 100 trees, respectively. It also suggests that most of the frequency distribution of basal area class are trimodal rather than bimodal under density stress. It is proposed that the leptokurtic distribution appears before normal distribution rather than direct change from platykurtic to normal distribution of basal area for selected stages in the development of stands.

• Korean Journal of Environment and Ecology
• /
• v.26 no.4
• /
• pp.593-603
• /
• 2012

This study was carried out to investigate the ecological succession sere and conservative value, and to provide the basic data for the planning of the Provincial Park Management in Seonamsagol(Valley), Jogyesan(Mt.) Provincial Park(altitude 884m), Suncheon City, Korea by analysing the structure of the plant community. Twenty plots(size is $20m{\times}20m$) were set up at an altitude of range from 315m to 480m. As a result of analysis of TWINSPAN which is one of the ordination technique, the plant communities were divided into four groups which are community I(Quercus variabilis community), community II(Q. serrata community), community III(Decideous broad-leaved plant community), and community IV(Carpinus tschonoskii community). The warmth index is $104^{\circ}C{\cdot}month$ based on the data of monthly mean temperature during the past thirty years(1981~2010), so we found out that the vegetation of the study site located in the South Temperate Climate Zone. We supposed that the ecological succession sere of the study site is in the early stage of developing from Q. serrata community to Carpinus tshonoskii community, however we should do a long-term monitoring to investigate the changes of the ecological succession each plant community, meanwhile Sasa borealis was dominant species in the shrub layer. The diameter at breast height of specimen tree is range from 20 to 55cm(average 36cm) and the height of that is range from 14 to 35m(average 23cm). The age of community I was 64 years old, that of community II was from 59 to 64 years old, that of community III was from 51 to 62 years old, and that of community IV was from 41 to 68 years old, thus the age of the study site is about from 38 to 72 years old. According to the index of Shnnon's diversity(unit: $400m^2$), community IV was ranged from 0.8452 to 1.2312, community III was ranged from 0.8044 to 1.1404, community II was ranged from 0.8221 to 0.9971, and community I was 0.8324.

• Journal of Wetlands Research
• /
• v.17 no.2
• /
• pp.124-129
• /
• 2015

In order to get as ecological basic data for river restoration, vegetation investigation was conducted in natural river and analysed it synecological methods, such as ordination cluster. 29 plant communities units were identified and the major dominant plant communites were Quercus mongolica community, Pinus densiflora community, Populus davidiana community, Q. variabilis community and Prunus sargentii community. River vegetations were classified into ravine and gorge forest type and riverine softwood forest type. Ravine and gorge forest was dominanted by hardwood which located in steep slope and in high elevation, and riverine softwood forest by softwood, salix spp. Naturality was an important criterion for the selection of rivers, so many of the selected rivers are located in the upper stream and mid stream rather than the lower stream, where more human intervention is involved. Plant communities were consisted of hardwood forest(44 plots, 92%) and softwood forest(4 plot, 8%), respectively. PCA with total layer data showed 5 groups of communities: Q. mongolica community group, Prunus sargentii community group, Pinus densiflora community group, Prunus sargentii community - Pinus densiflora community group and the rest communities group. PCA with tree layer showed 3 groups: Q. mongolica community group, Prunus sargentii community group, and the rest community group. Cluster analysis also a showed a similar communities group to PCA ordination, but Magnolia sieboldii community and Prunus sargentii community were distinguished from the PCA result. From the result, it can be concluded that the plant communities of riparian be divided into hardwood and softwood forest by statistical techniques. It was appropriate to plant species such as Quercus mongolica, Pinus densiflora, Populus davidiana, Quercus variabilis and Prunus sargentii, at levee zone and high water level. And Sliax spp. were appropriate for planted plants at waterfront and low water level. The herb species to be planted on the floodplain were recommanded in the species composition co-occurred with the woody species.

• Korean Journal of Heritage: History & Science
• /
• v.43 no.1
• /
• pp.260-279
• /
• 2010

The analysis on bibliography and field investigation of Callipogon relictus Semenov, 1898 (Korean natural monument number 218) shows that the size varies from country to country, and Korean specimens, for which male is 85~120mm and female is 65~85mm, are found to be the largest. The average diameter and length of egg are 2.60mm and 6.72mm respectively. The larva has milky color and is about 100~150mm in length. The pupa is nearly 70~110mm. An adult generally appears from June to September in Korea in the broadleaf forest of lowland, whereas it appears from June to July in China. It is known that the pupa largely feed on the old tree trunk of Carpinus laxiflora (Siebold & Zucc.) blume in Korea, but no such data have been reported in China and Russia, showing differences in host plants. While the larva period is not exactly known in Korea, it is reported to be two years in China. It appears that the species inhabits in very limited regions of approximately between geographical latitude $37.5^{\circ}{\sim}47.8^{\circ}$ and longitude $126^{\circ}{\sim}140^{\circ}$ including Korea, China and Russia. To conserve the long-horned beetle in Korea, this research drew out following some conclusions through analyzing the references and field survey data. First, it need to perform precise survey on the natural environment of occurring and collected area or place including host plant kinds, temperate, humidity, latitude, longitude etc. Second, habitat region must be designated as a restricted development area, and it need to exclude or reduce the damage factors to prosper reproduction of the species. Third, it is necessary to keep loosing cautiously artificial breeding individuals in the reported sites, not disturbing scope of natural populations. Fourth, it needs to educate or publicize many people importance and value of this species through many methods.

• Journal of Korea Water Resources Association
• /
• v.56 no.5
• /
• pp.311-323
• /
• 2023

In order to reduce disaster damage by localized heavy rains, floods, and urban inundation, it is important to know in advance whether natural disasters occur. Currently, heavy rain watch and heavy rain warning by the criteria of the Korea Meteorological Administration are being issued in Korea. However, since this one criterion is applied to the whole country, we can not clearly recognize heavy rain damage for a specific region in advance. Therefore, in this paper, we tried to reset the current criteria for a special weather report which considers the regional characteristics and to predict the damage caused by rainfall after 1 hour. The study area was selected as Gyeonggi-province, where has more frequent heavy rain damage than other regions. Then, the rainfall inducing disaster or hazard-triggering rainfall was set by utilizing hourly rainfall and heavy rain damage data, considering the local characteristics. The heavy rain damage prediction model was developed by a decision tree model and a random forest model, which are machine learning technique and by rainfall inducing disaster and rainfall data. In addition, long short-term memory and deep neural network models were used for predicting rainfall after 1 hour. The predicted rainfall by a developed prediction model was applied to the trained classification model and we predicted whether the rain damage after 1 hour will be occurred or not and we called this as 1ST-Model. The 1ST-Model can be used for preventing and preparing heavy rain disaster and it is judged to be of great contribution in reducing damage caused by heavy rain.