• Journal of Intelligence and Information Systems
• /
• v.23 no.2
• /
• pp.107-122
• /
• 2017

Volatility in the stock market returns is a measure of investment risk. It plays a central role in portfolio optimization, asset pricing and risk management as well as most theoretical financial models. Engle(1982) presented a pioneering paper on the stock market volatility that explains the time-variant characteristics embedded in the stock market return volatility. His model, Autoregressive Conditional Heteroscedasticity (ARCH), was generalized by Bollerslev(1986) as GARCH models. Empirical studies have shown that GARCH models describes well the fat-tailed return distributions and volatility clustering phenomenon appearing in stock prices. The parameters of the GARCH models are generally estimated by the maximum likelihood estimation (MLE) based on the standard normal density. But, since 1987 Black Monday, the stock market prices have become very complex and shown a lot of noisy terms. Recent studies start to apply artificial intelligent approach in estimating the GARCH parameters as a substitute for the MLE. The paper presents SVR-based GARCH process and compares with MLE-based GARCH process to estimate the parameters of GARCH models which are known to well forecast stock market volatility. Kernel functions used in SVR estimation process are linear, polynomial and radial. We analyzed the suggested models with KOSPI 200 Index. This index is constituted by 200 blue chip stocks listed in the Korea Exchange. We sampled KOSPI 200 daily closing values from 2010 to 2015. Sample observations are 1487 days. We used 1187 days to train the suggested GARCH models and the remaining 300 days were used as testing data. First, symmetric and asymmetric GARCH models are estimated by MLE. We forecasted KOSPI 200 Index return volatility and the statistical metric MSE shows better results for the asymmetric GARCH models such as E-GARCH or GJR-GARCH. This is consistent with the documented non-normal return distribution characteristics with fat-tail and leptokurtosis. Compared with MLE estimation process, SVR-based GARCH models outperform the MLE methodology in KOSPI 200 Index return volatility forecasting. Polynomial kernel function shows exceptionally lower forecasting accuracy. We suggested Intelligent Volatility Trading System (IVTS) that utilizes the forecasted volatility results. IVTS entry rules are as follows. If forecasted tomorrow volatility will increase then buy volatility today. If forecasted tomorrow volatility will decrease then sell volatility today. If forecasted volatility direction does not change we hold the existing buy or sell positions. IVTS is assumed to buy and sell historical volatility values. This is somewhat unreal because we cannot trade historical volatility values themselves. But our simulation results are meaningful since the Korea Exchange introduced volatility futures contract that traders can trade since November 2014. The trading systems with SVR-based GARCH models show higher returns than MLE-based GARCH in the testing period. And trading profitable percentages of MLE-based GARCH IVTS models range from 47.5% to 50.0%, trading profitable percentages of SVR-based GARCH IVTS models range from 51.8% to 59.7%. MLE-based symmetric S-GARCH shows +150.2% return and SVR-based symmetric S-GARCH shows +526.4% return. MLE-based asymmetric E-GARCH shows -72% return and SVR-based asymmetric E-GARCH shows +245.6% return. MLE-based asymmetric GJR-GARCH shows -98.7% return and SVR-based asymmetric GJR-GARCH shows +126.3% return. Linear kernel function shows higher trading returns than radial kernel function. Best performance of SVR-based IVTS is +526.4% and that of MLE-based IVTS is +150.2%. SVR-based GARCH IVTS shows higher trading frequency. This study has some limitations. Our models are solely based on SVR. Other artificial intelligence models are needed to search for better performance. We do not consider costs incurred in the trading process including brokerage commissions and slippage costs. IVTS trading performance is unreal since we use historical volatility values as trading objects. The exact forecasting of stock market volatility is essential in the real trading as well as asset pricing models. Further studies on other machine learning-based GARCH models can give better information for the stock market investors.

• The korean journal of orthodontics
• /
• v.29 no.4 s.75
• /
• pp.467-481
• /
• 1999

Cleft lip and/or palate is one of the most common congenital craniofacial anomalies. According to previous epidemiologic studies, incidence of cleft lip and/or palate has been increasing nowadays. However, there is no report about epidemiologic study of cleft lip and/or palate patients who visited dept. of orthodontics in Korea. So the purpose of this study was to provide the epidemiological characteristics and important basic clinical data for the diagnosis and the treatment of the cleft lip and/or palate patients. With the orthodontic and cleft charts, diagnostic models and X-ray films from 250 patients with cleft lip and/or palate who visited Dept. of Orthodontics, Seoul National University Dental Hospital during the last 11 years, the authors investigated patient's visiting yew, types of cleft, patient's gender, and Angle's classification of malocclusion, and surgery timing. The results were as follows ; 1. The number of cleft patients who visited Dept. of Orthodontics, SNUDH increased during 1988-1990 and then it declined until 1992. From 1993 to 1996, it showed a stationary trend. After 1997 it showed an overwhelmingly increasing trend. 2. In the cleft type, the ratio of cleft lip cleft lip and alveolus cleft palate : cleft lip and palate was 7.6:19.2:9.6:63.6. In cleft position, unilateral clefts were more than bilateral ones (cleft lip 79:21, cleft lip and alveolus 77:23, cleft lip and palate 75.5:24.5). In cleft side, left clefts were mote than right clefts (cleft lip 53.3:46.7 cleft lip and alveolus 59.5:40.5, cleft lip and palate 59.2:40.8). 3. In gender ratio, males were more than females in cleft lip (57.9:42.1), cleft lip and alveolus (68.8:31.2) and cleft lip and palate (76.1:23.9). But in cleft Palate females were more than males as 41.7: 58.3. 4. In the age groups, 7-12 year group was the most abundant as $52\%$, and then 0-6 year group ($20.4\%$), 13-18 year group ($17.2\%$), more than 18 yew group ($10.4\%$) were followed as descending order. 5. Most of the cleft lip repair surgeries were operated in 0-3 month ($60.3\%$) and 4-6 month ($17.9\%$). 6. The cleft palate repair surgeries were done in 1-2 year ($31.7\%$), 0-1 year ($25.6\%$), 2-3 year ($12.1\%$), more than 5 year ($11.6\%$) as descending order. 7. The lip scar revision surgeries were done before admission at elementary school in $60\%$. (4-6 you ($27.5\%$), 6-8 year ($19.6\%$), more than 10 year ($19.6\%$), 2-4 year ($13.7\%$) as descending order) 8. The rhinoplasties were done before admission at elementary school in $51.7\%$. (0-2 year ($7.1\%$), 2-4 year ($14.3\%$), 4-6 year ($21.4\%$), 6-8 year ($14.3\%$)). 9. The pharyngeal flap were done at 6 Y (72.5 months) after birth on average and there was even distribution of surgery timing. 10. In relationship between Angle's classification of malocclusion and cleft types, Class I was most abundant and Class III, Class II were followed as descending order in cleft lip group. But Class III was most abundant and Class I, Class II were followed as descending order in cleft lip and alveolus group, cleft palate group, and cleft lip and Palate group. The percentage of frequency in Class III malocclusion was overwhelmingly higher in cleft lip and palate group than any other groups. 11. Because the frequency of class III malocclusion was most prevalent in all age groups, anterior crossbite was the most common chief complaint of cleft patients.

• Journal of Korean Society of Forest Science
• /
• v.42 no.1
• /
• pp.1-38
• /
• 1979

The woody species of Genus Ilex which are endemic to Korea are distributed on limited area due to solely temperature factor. There is some differences according to species, however in general, the evergreen Ilex are found along southern coastal area of Korean Peninsula and near islands where the cold index does not exceed $-5^{\circ}C$. But Ilex macropoda and the variety, only deciduous ones, are grown in temperate zone of the peninsula and some islands. The list of Ilex species of Korea are as follows. Ilex cornuta Lindley et Pax., I. crenata Thunb. var. microphylla Max., I. crenata Thunb., I. rotunda Thunb., I. macropoda Miq., I. macropoda Miq. var. pseudo-macropoda Loensner, I. integra Thunb. The author surveyed the populations of Ilex species as many as possible and data of some characters such as leaf shape, spine, fruit shape, stomata density, sex ratio in natural communities, etc. are collected. Almost all the Ilex species in Korea show sporadic distribution. This means quite small sized populations isolate distantly each other eliminating the change of gene exchange in between. Particularly Ilex conuta and I. crenata show the morphological differentiation among populations as well as significant individual variation within a population. These were true with such characteristics, leaf shape, leaf dimension, leaf margin, fruit shape, spine, and stomata density. The founded are that the fruit length and the stomata density counted on the beneath surface of leaves of Ilex cornuta increased with the decrease of latitude. These are naturally closely related with the cold index values. The table shown below indicates the correlation between mean stomata density per $0.3642mm^2$ and cold index values. These relation however were not observed on Ilex crenata. The most dominated natured in relation to individual variation were outline of leaf, the number of marginal spine, the shape of leaf cross section and the degree of luster of the upper leaf surface. As shown in photos 5~7, these variations are agreed at a glance. There are reports that the development of marginal spines in some Ilex species is associated with the juvenility and topophysis. In present study, these two factors were neglected because of the intended sampling procedure. Of Ilex rotunda, population difference with the characteristics of leaf length is recognized but not for leaf width, petiole length, and fruit size. However, individual variations within a population were significantly large. In case of Ilex integra, only individual differences within population were calculated statistically for such characteristics as leaf length, leaf width, and petiole length. As to natural population, the sex ratio was 1:2 (female to male) for Ilex cornuta, and 1:1 for Ilex crenata. The tendency of more male than female in I. cornuta was agreed to other observations. Preparing the tip cutting of length 10cm, and treating with IBA, then attaching earth ball to the cut end, very successful rooting percentages were obtained. Asexual propagation has the advantages of maintaining the heterozygosity of existing varieties and overcoming the difficulties of delayed seed germination frequently encountered with Ilex species. Considering a great deal of variation in morphological traits, a good possibility of selection breeding for decorative and ornamental purposes exists. At present, these evergreen Ilex are ignored by local people as nuisance weedy shrubs. So the proper protection measures should promptly be taken.

• Tuberculosis and Respiratory Diseases
• /
• v.46 no.5
• /
• pp.674-684
• /
• 1999

Background : It has been generally known that the incidence of lung cancer is higher in the patients with idopathic pumonary fibrosis (IPF) than those in general population. The reported incidence was variable from 4.8 to 43.2%. There were controversies on the most frequent cell type (squamous cell carcinoma vs. adenocarcinoma) and no study was done about the real concordance of cancer and the fibrotic lesion. And the pulmonary fibrosis may influence not only the development of cancer but also the treatment and prognosis of the cancer, but there was no report on that point. Method : Total 63 patients ($66.8{\pm}7.8$ year, M : F=61 : 2) were diagnosed as IPF combined with lung cancer (IFF-CA) at Asan Medical Center. A retrospective analysis was done about the risk factors of the lung cancer, pulmonary function test, the site of cancer(especially the relationship of the cancer with the fibrotic lesion), the histologic types, and the stage of cancer. The histologic types were compared with those of 2,660 patients with lung cancer who were diagnosed at the same institute for the same period. The effect of IPF on the treatment of the cancer was evaluated with the survival time after the detection of lung cancer. Results : The lung cancer was found in 63(22.9%) out of 281 patients with IPF. But in most of them(45 patients), lung cancer was detected at the same time with IPF and only in 18 patients, the cancer was diagnosed during the follow-up($25.2{\pm}17.7$ months) of IPF. So in our study, 6.7% of patients with IPF developed lung cancer during the course of the disease. The age ($66.8{\pm}7.84$ vs. $63.4{\pm}11.1$ years), percentage of smoker (88.9 vs. 67.2%), and the male gender (96.8 vs. 67.6%) were significantly higher in IPF-CA compared with lone IPF (p<0.05). The odds ratio of smoking was 4.7 compared with non smoking IPF controls. The lung cancer was located more frequently in the upper lobe and 55.5% was in the periphery of lung. The cancer was developed in the fibrotic lesion in 23 patients (35.9%), and in the majority of the patients, the cancer was separated from the fibrosis. The cell type of the lung cancer in IPF-CA was squamous cell carcinoma 34.9%, adenocarcinoma 30.2%, small cell carcinoma 19.0%, large cell undifferenciated carcinoma 6.3%, and others 9.5%. No significant difference in the distribution of histologic type of the lung cancer was found between IPF-CA and lone lung cancer. There was no significant difference in demographic features, cell types, location and the stage of the cancer between the group with concurrent IPF-CA and the group with cancer diagnosed during the follow up of IPF. There was a tendency (but statistically not significant : p=0.081) of higher incidence of adenocarcinoma among the cancers developed in the fibrotic area(43.5%) (F-CA) than in the cancers in non-fibrotic area (22.5%) (NF-CA). The prognosis of the patients with F-CA was poor (median survival : 4 months) compared with the patients with NF-CA (7 months, p=0.013), partly because the prevalence of severe IPF (the extent of fibrosis in HRCT 50%) was higher in F-CA group. Conclusion : These data suggest that the lung cancer in the patients with IPF has similar features to the ordinary lung cancer.