• ETRI Journal
• /
• v.45 no.6
• /
• pp.1007-1021
• /
• 2023

Stroke is the leading cause of permanent disability in adults, and it can cause permanent brain damage. According to the World Health Organization, 795 000 Americans experience a new or recurrent stroke each year. Early detection of medical disorders, for example, strokes, can minimize the disabling effects. Thus, in this paper, we consider various risk factors that contribute to the occurrence of stoke and machine learning algorithms, for example, the decision tree, random forest, and naive Bayes algorithms, on patient characteristics survey data to achieve high prediction accuracy. We also consider the semisupervised self-training technique to predict the risk of stroke. We then consider the near-miss undersampling technique, which can select only instances in larger classes with the smaller class instances. Experimental results demonstrate that the proposed method obtains an accuracy of approximately 98.83% at low cost, which is significantly higher and more reliable compared with the compared techniques.

• Asia-Pacific Journal of Business
• /
• v.15 no.1
• /
• pp.145-157
• /
• 2024

Purpose - In this study, we propose an empirical model for predicting seasoned equity offering (SEO here after) using machine learning methods. Design/methodology/approach - The models utilize the random forest method based on decision trees that considers non-linear relationships, as well as the gradient boosting tree model. SEOs incur significant direct and indirect costs. Therefore, CEOs' decisions of seasoned equity issuances are made only when the benefits outweigh the costs, which leads to a non-linear relationship between SEOs and a determinant of them. Particularly, a variable related to market timing effectively exhibit such non-linear relations. Findings - To account for these non-linear relationships, we hypothesize that decision tree-based random forest and gradient boosting tree models are more suitable than the linear methodologies due to the non-linear relations. The results of this study support this hypothesis. Research implications or Originality - We expect that our findings can provide meaningful information to investors and policy makers by classifying companies to undergo SEOs.

• Journal of Korean Society of Forest Science
• /
• v.105 no.3
• /
• pp.360-365
• /
• 2016

The purpose of this study is to provide the basic information for the early enforcement and extension of the improvement project of management scale of private forest land by understanding the characteristics of forest owners, who have an influence on the participation of the project as one of the private forest management vitalization plans. To achieve this goal, a questionnaire survey targeting 373 forest owners was conducted and analyzed by Binary-Logistic Regression. The variables for binary-logistic regression included gender, age, academic ability, occupation, income, residence, purpose of forest ownership, and status of cooperative membership. As a result of the analysis, 267 forest owners (71.6%) of total 373 forest owners have the intention to participate in the scaling project for private forest management. The rest of forest owners (106 forest owners, 28.4%) would not be willing to participate in the project. As a result of binary-logistic regression, the most important variables, which have an impact on the participation of private forest management scale improvement project, are age, job and forest own purpose.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.25 no.5
• /
• pp.43-57
• /
• 2022

This study aims to classify forest fire-affected areas, identify forest types by the intensity of forest fire damage using multi-time Landsat-satellite images before and after forest fires and to analyze the effects of artificial restoration sites and natural restoration sites. The difference in the values of the Normalized Burned Ratio(NBR) before and after forest fire damage not only maximized the identification of forest fire affected and unaffected areas, but also quantified the intensity of forest fire damage. The index was also used to confirm that the higher the intensity of forest fire damage in all forest fire-affected areas, the higher the proportion of coniferous forests, relatively. Monitoring was conducted after forest fires through Normalized Difference Vegetation Index(NDVI), an index suitable for the analysis of effects by restoration type and the NDVI values for artificial restoration sites were found to no longer be higher after recovering the average NDVI prior to the forest fire. On the other hand, the natural restoration site witnessed that the average NDVI value gradually became higher than before the forest fires. The study result confirms the natural resilience of forests and these results can serve as a basis for decision-making for future restoration plans for the forest fire affected areas. Further analysis with various conditions is required to improve accuracy and utilization for the policies, in particular, spatial analysis through forest maps as well as review through site checks before and immediately after forest fires. More precise analysis on the effects of restoration will be available based on a long term monitoring.

• Journal of Climate Change Research
• /
• v.5 no.1
• /
• pp.21-36
• /
• 2014

Forest ecosystem provides variety goods and services for human being. Unlike goods, forest ecosystem services could not be easily priced by market mechanism. This uncertainty has been caused to conflict in decision-making related forest ecosystem services. Quantification of forest ecosystem services is required to understand the importance of ecosystem services and their contribution to decision-making. As a growing concern of climate change, it is necessary to quantify and calculate carbon storage and sequestration in forest. In this study, for quantifying carbon storage and sequestration, we compared scale, output, input data availability of the models and analyzed the applicability of the models to Korea. The results of this study show that most models are applicable for quantifying carbon storage and sequestration. However, relatively few models are applicable for other regulating services (air quality regulation, flood mitigation, erosion control, water quality, etc.) of forest. This study would be helpful for quantifying regulating services of forest ecosystem research.

• Journal of Korean Society of Forest Science
• /
• v.102 no.3
• /
• pp.415-419
• /
• 2013

To mitigate forest fire damage, it is needed to concentrate suppression resources on the fire having a high probability to become large in the initial stage. The objective of this study is to develop the large fire judgement model which can estimate large fire possibility index between the fire size and the related factors such as weather, terrain, and fuel. The results of logistic regression equation indicated that temperature, wind speed, continuous drought days, slope variance, forest area were related to the large fire possibility positively but elevation has negative relationship. This model may help decision-making about size of suppression resources, local residents evacuation and suppression priority.

• Journal of Korean Society of Forest Science
• /
• v.76 no.4
• /
• pp.390-396
• /
• 1987

The model and its example were provided to predict wood production for a district forest planning. The method of Gentan probability is widely accepted for the prediction of wood production. The suggested model is different in the decision of cutting age distribution from that of Prof. Suzuki; the former can use either Weibull distribution or Gamma distribution, but the latter is possible only by Gamma distribution. This developed system can be used not only for establishing a district forest planning, but also for providing forest management information, such as periodic harvest volume, growing stock, labor requirement, and so forth.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38A no.1
• /
• pp.1-9
• /
• 2013

This paper propose a human action recognition method that uses bag-of-features (BoF) based on CS-LBP (center-symmetric local binary pattern) and a spatial pyramid in addition to the random forest classifier. To construct the BoF, an image divided into dense regular grids and extract from each patch. A code word which is a visual vocabulary, is formed by k-means clustering of a random subset of patches. For enhanced action discrimination, local BoF histogram from three subdivided levels of a spatial pyramid is estimated, and a weighted BoF histogram is generated by concatenating the local histograms. For action classification, a random forest, which is an ensemble of decision trees, is built to model the distribution of each action class. The random forest combined with the weighted BoF histogram is successfully applied to Standford Action 40 including various human action images, and its classification performance is better than that of other methods. Furthermore, the proposed method allows action recognition to be performed in near real-time.

• Journal of Digital Convergence
• /
• v.19 no.12
• /
• pp.477-486
• /
• 2021

This study was conducted to detect metabolic syndrome risk factors and gender difference in adults. 18,616 cases of adults are collected by Korea Health and Nutrition Examination Study from 2016 to 2019. Using 4 types of machine Learning(Logistic Regression, Decision Tree, Naïve Bayes, Random Forest) to predict Metabolic Syndrome. The results showed that the Random Forest was superior to other methods in men and women. In both of participants, BMI, diet(fat, vitamin C, vitamin A, protein, energy intake), number of underlying chronic disease and age were the upper importance. In women, education level, menarche age, menopause was additional upper importance and age, number of underlying chronic disease were more powerful importance than men. Future study have to verify various strategy to prevent metabolic syndrome.

• Journal of Broadcast Engineering
• /
• v.23 no.1
• /
• pp.115-125
• /
• 2018

This paper presents a fast partition decision framework for High Efficiency Video Coding (HEVC) Screen Content Coding (SCC) based on machine learning. Currently, the HEVC performs quad-tree block partitioning process to achieve optimal coding efficiency. Since this process requires a high computational complexity of the encoding device, the fast encoding process has been studied as determining the block structure early. However, in the case of the screen content video coding, it is difficult to apply the conventional early partition decision method because it shows different partition characteristics from natural content. The proposed method solves the problem by classifying the screen content blocks after partition decision, and it shows an increase of 3.11% BD-BR and 42% time reduction compared to the SCC common test condition.