• Journal of the Korea Society of Computer and Information
• /
• v.27 no.11
• /
• pp.147-155
• /
• 2022

In this study, we developed a system to dynamically balance a daily stock portfolio and performed trading simulations using gradient boosting and genetic algorithms. We collected various stock market data from stocks listed on the KOSPI and KOSDAQ markets, including investor-specific transaction data. Subsequently, we indexed the data as a preprocessing step, and used feature engineering to modify and generate variables for training. First, we experimentally compared the performance of three popular gradient boosting algorithms in terms of accuracy, precision, recall, and F1-score, including XGBoost, LightGBM, and CatBoost. Based on the results, in a second experiment, we used a LightGBM model trained on the collected data along with genetic algorithms to predict and select stocks with a high daily probability of profit. We also conducted simulations of trading during the period of the testing data to analyze the performance of the proposed approach compared with the KOSPI and KOSDAQ indices in terms of the CAGR (Compound Annual Growth Rate), MDD (Maximum Draw Down), Sharpe ratio, and volatility. The results showed that the proposed strategies outperformed those employed by the Korean stock market in terms of all performance metrics. Moreover, our proposed LightGBM model with a genetic algorithm exhibited competitive performance in predicting stock price movements.

• Journal of Technology Innovation
• /
• v.30 no.1
• /
• pp.57-104
• /
• 2022

The government has supported the innovation of private firms by intervening the market for various purposes, such as preventing market failure, alleviating information asymmetry, and allocating resources efficiently. Although the government's R&D budget increased rapidly in the 2000s, it is not clear whether the government intervention has made desirable impact on the market. To address this, the current study attempts to explore this issue by doing a systematic literature review on foreign and domestic papers in an integrated way. In total, 168 studies are analyzed using contents analysis approach and various lens, such as policy additionality, policy tools, firm size, unit of analysis, data and method, are adopted for analysis. Overlapping policy target, time lag between government intervention and policy effects, non-linearity of financial supports, interference between different polices, and out-dated R&D tax incentive system are reported as factors hampering the effect of the government intervention. Many policy prescriptions, such as program evaluation indices reflecting behavioral additionality, an introduction of policy mix and evidence-based policy using machine learning, are suggested to improve these hurdles.

• Land and Housing Review
• /
• v.15 no.3
• /
• pp.113-127
• /
• 2024

This study explores shifts in real estate investment sentiment using media reports from 2012 to 2022, segmenting the market dynamics into three distinct cycles based on housing and land transaction indices. Leveraging 54 BigKinds media sources, we investigates 3,387 headlines and 8,544 body texts using LDA topic modeling. The results show that the first cycle (2012-2015 ) centered on apartment pre-sales, where policy changes influenced sentiment but did not consistently affect investment decisions. The second cycle (2016-2018) was characterized by interest rate hikes and rising property prices in Seoul, resulting in significant fluctuations in transaction volumes. The third cycle (2019-2022) encompassed the effects of COVID-19, market instability, and policy failures, leading to distorted and weakened investment sentiment. Each cycle demonstrated that policies, interest rates, and economic events significantly shaped investor sentiment, as reflected in media reports.

• Communications for Statistical Applications and Methods
• /
• v.16 no.4
• /
• pp.627-637
• /
• 2009

Our economic system highly depends on the international trade market and demands high level of foreign language abilities to the young generation. Scores of various internationally authorized language tests are used to estimate students's language ability before giving their admission to universities or high schools. According to increase of unemployment rate of the young people, universities put their strength on the language abilities of their students, and each university uses some authorized language scores before determining students who win scholarships. This study suggests two types of indices to evaluate students's relative achievement of foreign language abilities between two time points considering their current abilities. One, named one-side relative achievement index, is defined as the ratio of additional score between two time points to the remaining score at the base time point to the full score. The second one, named two-side relative achievement index, is defined as the same manner as the first one if the score is improved, and if not, is defined as the ratio of the amount of losing scores to the current score at the base time point. Two-side achievement index is more useful since it has smaller variation than the former and is easier to interpret. However, both indices are useful to compare the achievements of different tests.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.21 no.1
• /
• pp.209-220
• /
• 2022

With the worldwide spread of African swine fever, interest in livestock epidemics has increased. Livestock transport vehicles are the main cause of the spread of livestock epidemics, but there are no empirical quarantine procedures and standards related to the mobility of livestock transport vehicles in South Korea. This study extracted the trajectory of livestock-related vehicles using the facility-visit history data from the Korea Animal Health Integrated System and the DTG (Digital Tachograph) data from the Korea Transportation Safety Authority. The results are presented as exposure indices aggregating the link-time occupancy of each vehicle. As a result, 274,519 livestock-related vehicle trajectories were extracted, and the exposure values by link and zone were derived quantitatively. This study highlights the need for prior monitoring of livestock transport vehicles and the establishment of post-disaster prevention policies.

• Asian-Australasian Journal of Animal Sciences
• /
• v.26 no.2
• /
• pp.151-156
• /
• 2013

Hanwoo (Also known as Korean native cattle; Bos taurus coreanae) have been used for transportation and farming for a long time in South Korea. It has been about 30 yrs since Hanwoo improvement began in earnest as beef cattle for meat yield. The purpose of this study was to determine the trend of improvement as well as to estimate genetic parameters of the traits being used for seedstock selection based on the data collected from the past. Hanwoo proven bulls in South Korea are currently selected through performance and progeny tests. National Hanwoo genetic evaluations are implemented with yearling weight (YW), carcass weight (CW), eye muscle area (EMA), backfat thickness (BF) and marbling score (MS). Yearling weights and MS are used for selecting young bulls, and EMA, BF, and MS are used for selecting proven bulls. One individual per testing room was used for performance tests, and five individuals per room for progeny tests. Individuals tested were not allowed to graze pasture, but there was enough space for them to move around in the testing room. Feeds including roughages and minerals were fed ad libitum, and concentrates were provided at the rate of about 1.8% of individual weight. Overall means of the traits were $352.8{\pm}38.56$ kg, $335.09{\pm}44.61$ kg, $77.85{\pm}8.838\;cm^2$, $8.6{\pm}3.7$ mm and $3.293{\pm}1.648$ for YW, CW, EMA, BF and MS. Heritabilities estimated in this study were 0.30, 0.30, 0.42, 0.50 and 0.63 in YW, CW, EMA, BF and MS, respectively, which are similar to results from previous research. Yearling weight was 315.54 kg in 1998, and had increased to 355.06 kg in 2011, resulting in about 40 kg of improvement over 13 yrs. YW and CW have improved remarkably over the past 15 yrs. Breeding values between 1996 and 2000 decreased or did not change much, but have moved in a desirable direction since 2001. These improvements correspond with the substantial increase in use of animal models since the late 1990s in Korea. Hanwoo testing programs have practically contributed to the improvement in aspects of quality and quantity. In sum, the current selection system is good enough to accommodate circumstances where fewer sires are used on many more cows. Although progeny tests take longer and cost more, they seem to be appropriate under the circumstances of the domestic market with its higher requirement for better meat quality. Consequently, accumulative data collection, genetic evaluation model development, revision of selection indices, as well as cooperation among farms, associations, National Agricultural Cooperative Federation, universities, research institutes, and government agencies must be applied to the Hanwoo selection program. All these efforts will assist the domestic market to secure a competitive position against imported beef under Free Trade Agreement trade system and will provide farmers with higher profits as well as the public with a higher quality of beef.

• The Korean Journal of Financial Management
• /
• v.20 no.2
• /
• pp.151-180
• /
• 2003

We examine the interdependence of the major Asia-Pacific stock markets including S&P 500, FTSE 100, Kualar Lumpur Composite, Straits Times, Hang Seng, NIKKEI 225 and KOSPI 200 from October 4, 1995 to March 31,2000. The analysis employs the vector-auto-regression, Granger causality, impulse response function and variance decomposition using daily returns on the national stock market indices. The findings in this paper indicate that the volatilities of all countries has grown after IMF crisis, while there is no significance in cointegration test of both total period and sub-periods. This result implies that investors are able to get abnormal returns by investment diversification according to the portfolio theory. We find that while the effect from NIKKEI 225 to others is relatively weak, the interdependence from S&P 500 to other countries is strong. Also we find that the strong effect from Straits Times to Hang Seng exists. This study suggests that there is slight feedback relation between KOSPI 200 and Kualar Lumpur Composite, Straits Times, Hang Seng stock market.

• Korea Real Estate Review
• /
• v.28 no.4
• /
• pp.63-73
• /
• 2018

The purpose of this study is to identify the key factors affecting official land values and the appropriateness of the assessed land price, to find out what determines the real estate price and to assess the appropriateness of the valuation. This study explored whether actual transaction prices of forest land located in six Gu districts in southern and northern parts of Han river are appropriate using independent sample t-analysis and logistic regression analysis. Results showed that regional differences and shape were adequate for development restriction areas, whether biotope was designated, whether or not to be preserved, differences in pitch, and differences in use, and differences in bearing and approach. Thorough analysis of unique factors that determine forest land prices must be carried out in advance and the findings should be applied to the examination and assessment of official land values. The forest land appraisal system is closely related to the public's economic activity, thus it is necessary to apply forest land value determinants considered to be significant by market participants to the forestland appraisal system. I look forward to seeing variables related to the appropriateness of forest land transactions drawn from this study being used as indices for settlement of forest land transaction orders and market stabilization.

• Environmental and Resource Economics Review
• /
• v.31 no.2
• /
• pp.233-260
• /
• 2022

In this paper, we analyzed how oil price fluctuations affect stock price by industry using the non-parametric quantile causality test method. We used weekly data of WTI spot price, KOSPI index, and 22 industrial stock indices from January 1998 to April 2021. The empirical results show that the effect of changes in oil prices on the KOSPI index was not significant, which can be attributed to mixed responses of diverse stock prices in several industries included in the KOSPI index. Looking at the stock price response to oil price by industry, the 9 of 18 industries, including Cloth, Paper, and Medicine show a causality with oil prices, while 9 industries, including Food, Chemical, and Non-metal do not show a causal relationship. Four industries including Medicine and Communication (0.45~0.85), Cloth (0.15~0.45), and Construction (0.5~0.6) show causality with oil prices more than three quantiles consecutively. However, the quantiles in which causality appeared were different for each industry. From the result, we find that the effects of oil price on the stock prices differ significantly by industry, and even in one industry, and the response to oil price changes is different depending on the market situation. This suggests that the government's macroeconomic policies, such as industrial and employment policies, should be performed in consideration of the differences in the effects of oil price fluctuations by industry and market conditions. It also shows that investors have to rebalance their portfolio by industry when oil prices fluctuate.

• 한국노년학
• /
• v.25 no.2
• /
• pp.73-93
• /
• 2005

This study explores the determinants of economic status and well-being among elderly Koreans. Particular emphasis is paid to the role of intergenerational support in these processes because institutional responses to an aging population in Korea are currently inadequate. Path model results indicate that status maintenance variables have significant effects on economic status and well-being as measured by comprehensive income. Further, inequality in comprehensive income - as measured by the Gini and Theil's indices - is substantial and increases in successive age groups. However, market-based and family-based income, two components of comprehensive income, show different patterns. Inequality in market-based income is apparently increasing, while inequality in family-based income is decreasing gradually in successive age groups. With respect to the role of intergenerational support, results indicate that family-income-focused persons possess highly vulnerable sociodemographic characteristics, and that for the socially disadvantaged elderly Koreans, intergenerational support plays a crucial role in the later stages of life. This study also discusses the implications of these findings for the transition from latency to manifestation of intergenerational support in Korea.