• Journal of Korean Society for Geospatial Information Science
• /
• v.20 no.3
• /
• pp.11-18
• /
• 2012

Increasing concern about biodiversity has lead to a rise in demand on the spatial assessment of biological resources such as biodiversity assessment, protected area selection, habitat management and restoration in Korea. The purpose of this study is to create species richness map through data collection and modeling techniques for wildlife habitat assessment. The GAM (Generalized Additive Model) is easy to interpret and shows better relationship between environmental variables and a response variable than an existing overlap analysis and GLM (Generalized Linear Model). The study area delineated by a large watershed contains Jirisan national park, Mt. Baekun and Sumjin river with three kinds of protected areas (a national park, a landscape ecology protected area and an otter protected area). We collected the presence-absence data for wildlife (mammals and birds) using a stratified random sampling based on a land cover in the study area and implemented natural and socio-environmental data affecting wildlife habitats. After doing a habitat use analysis and specifying significant factors for each species, we built habitat suitability models using a presence-absence model and created habitat suitability maps for each species. Biodiversity maps were generated by taxa and all species using habitat suitability maps. Significant factors affecting each species habitat were different according to their habitat selection. Although some species like a water deer or a great tit were distributed at the low elevation, most potential habitats for mammals and birds were found at the edge of a national park boundary or near a forest around the medium elevation of a mountain range. This study will be used for a basis on biodiversity assessment and proected area selection carried out by Ministry of Environment.

• Korean Journal of Environment and Ecology
• /
• v.34 no.6
• /
• pp.589-600
• /
• 2020

Effective conservation and management of protected areas require monitoring the settlement of invasive alien species and reducing their dispersion capacity. We simulated the potential distribution of invasive alien plant species (IAPS) using three representative species distribution models (Bioclim, GLM, and MaxEnt) based on the IAPS distribution in the forest genetic resource reserve (2,274ha) in Uljin-gun, Korea. We then selected the realistic and suitable species distribution model that reflects the local region and ecological management characteristics based on the simulation results. The simulation predicted the tendency of the IAPS distributed along the linear landscape elements, such as roads, and including some forest harvested area. The statistical comparison of the prediction and accuracy of each model tested in this study showed that the GLM and MaxEnt models generally had high performance and accuracy compared to the Bioclim model. The Bioclim model calculated the largest potential distribution area, followed by GLM and MaxEnt in that order. The Phenomenological review of the simulation results showed that the sample size more significantly affected the GLM and Bioclim models, while the MaxEnt model was the most consistent regardless of the sample size. The optimal model overall for predicting the distribution of IAPS among the three models was the MaxEnt model. The model selection approach based on detailed flora distribution data presented in this study is expected to be useful for efficiently managing the conservation areas and identifying the realistic and precise species distribution model reflecting local characteristics.

• Journal of the Korean Society of Safety
• /
• v.33 no.1
• /
• pp.142-147
• /
• 2018

This study aims to analyze and discuss the accidents based on the level of traffic culture (LOT). In pursuing the above, LOT are divided into three categories based on the standardized index of traffic culture. Also, this study focuses on developing the accident models using GLM (generalized linear model). The main results are as follows. First, the null hypotheses that the ratios of fatal and serious injured persons (FSI) are the same over categories are rejected. Second, as the common variables, the ratio of turn signal usage and elderly population are analysed to be impacted to the ratio of FSI. Third, the traffic culture indicators among 5 accident factors which give impact to 'high level' are judged to affect the reduction of FSI. Fourth, compared to other levels, the traffic law violations among 7 accident factors of 'medium level' are estimated to influence the increase of FSI. Finally, in 'low level', the increasing ratio of traffic culture index compared to that of previous year and the number of hospital beds per person are evaluated to be significant to reducing the ratio of FSI. This study can be expected to give some policy implications to regional traffic safety policy-making.

• Journal of the Korea Society of Computer and Information
• /
• v.21 no.4
• /
• pp.89-96
• /
• 2016

In this paper, we study the relationship between vacancy duration and information items of a job posting by using generalized linear models and a decision tree analysis w.r.t. the three factors such as company characteristics, employment conditions, and constraints. The results indicate that the employment conditions rather than company characteristics are more influential to the vacancy duration. These effects are presumed to be based on the complex relations between the decisions of the employers and the job seekers. And in this paper we suggest the need to provide personalized and profiled labor market information tailored for a quick decision to job seekers and employers. Policy implication is that since employer's decision affects the vacation duration, employers may had better to provide a comprehensive labour market information including supply and demand of the required skills in order to reduce the time for judgment on the cost-effectiveness.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.19 no.6
• /
• pp.19-30
• /
• 2016

The projection of climate-related range shift is critical information for conservation planning of Korean fir (Abies koreana E. H. Wilson). We first modeled the distribution of Korean fir under current climate condition using five single-model species distribution models (SDMs) and the pre-evaluation weighted ensemble method and then predicted the distributions under future climate conditions projected with HadGEM2-AO under four $CO_2$ emission scenarios, the Representative Concentration Pathways (RCP) 2.6, 4.5, 6.0 and 8.5. We also investigated the predictive uncertainty stemming from five individual algorithms and four $CO_2$ emission scenarios for better interpretation of SDM projections. Five individual algorithms were Generalized linear model (GLM), Generalized additive model (GAM), Multivariate adaptive regression splines (MARS), Generalized boosted model (GBM) and Random forest (RF). The results showed high variations of model performances among individual SDMs and the wide range of diverging predictions of future distributions of Korean fir in response to RCPs. The ensemble model presented the highest predictive accuracy (TSS = 0.97, AUC = 0.99) and predicted that the climate habitat suitability of Korean fir would increase under climate changes. Accordingly, the fir distribution could expand under future climate conditions. Increasing precipitation may account for increases in the distribution of Korean fir. Increasing precipitation compensates the negative effects of increasing temperature. However, the future distribution of Korean fir is also affected by other ecological processes, such as interactions with co-existing species, adaptation and dispersal limitation, and other environmental factors, such as extreme weather events and land-use changes. Therefore, we need further ecological research and to develop mechanistic and process-based distribution models for improving the predictive accuracy.

• Computers and Concrete
• /
• v.33 no.2
• /
• pp.137-145
• /
• 2024

The technique of experimentally determining concrete's compressive strength for a given mix design is time-consuming and difficult. The goal of the current work is to propose a best working predictive model based on different machine learning algorithms such as Gradient Boosting Machine (GBM), Stacked Ensemble (SE), Distributed Random Forest (DRF), Extremely Randomized Trees (XRT), Generalized Linear Model (GLM), and Deep Learning (DL) that can forecast the compressive strength of ternary geopolymer concrete mix without carrying out any experimental procedure. A geopolymer mix uses supplementary cementitious materials obtained as industrial by-products instead of cement. The input variables used for assessing the best machine learning algorithm not only include individual ingredient quantities, but molarity of the alkali activator and age of testing as well. Myriad statistical parameters used to measure the effectiveness of the models in forecasting the compressive strength of ternary geopolymer concrete mix, it has been found that GBM performs better than all other algorithms. A sensitivity analysis carried out towards the end of the study suggests that GBM model predicts results close to the experimental conditions with an accuracy between 95.6 % to 98.2 % for testing and training datasets.

• Korean Journal of Ichthyology
• /
• v.20 no.2
• /
• pp.105-111
• /
• 2008

A pelagic tuna longline research cruise in the eastern and central Pacific Ocean from September to October of 2006 was conducted to compare catch rates with the use of different hook type and bait combinations. Traditional tuna hooks (J 4) and three circle hook types (C15, C16, C18), along with five bait types (chub mackerel (CM), jack mackerel (JM), milkfish (MF), sardine (SD), and squid (SQ)) and hook number as a proxy for hook depth were evaluated for their effect on bigeye tuna catch rates (fish per 1,000 hooks) using Generalized Linear Models (GLMs). Results from 28 sets indicated significant differences in bigeye catch rates between individual longline sets and hook number. The GLM explained 33% of the deviance in bigeye catch rates with these two factors. An alternative model formulation included bait type which had a small effect (explaining 2.7% of the deviance) on catch rates. Hook type had a negligible and non-significant effect in the GLMs. These results indicate that all of the hooks and baits tested are equally effective at catching bigeye tuna and that hook number (depth) was the paramount operational factor in explaining bigeye tuna catch rates.

• Journal of Intelligence and Information Systems
• /
• v.23 no.3
• /
• pp.139-153
• /
• 2017

Corporate bankruptcy can cause great losses not only to stakeholders but also to many related sectors in society. Through the economic crises, bankruptcy have increased and bankruptcy prediction models have become more and more important. Therefore, corporate bankruptcy has been regarded as one of the major topics of research in business management. Also, many studies in the industry are in progress and important. Previous studies attempted to utilize various methodologies to improve the bankruptcy prediction accuracy and to resolve the overfitting problem, such as Multivariate Discriminant Analysis (MDA), Generalized Linear Model (GLM). These methods are based on statistics. Recently, researchers have used machine learning methodologies such as Support Vector Machine (SVM), Artificial Neural Network (ANN). Furthermore, fuzzy theory and genetic algorithms were used. Because of this change, many of bankruptcy models are developed. Also, performance has been improved. In general, the company's financial and accounting information will change over time. Likewise, the market situation also changes, so there are many difficulties in predicting bankruptcy only with information at a certain point in time. However, even though traditional research has problems that don't take into account the time effect, dynamic model has not been studied much. When we ignore the time effect, we get the biased results. So the static model may not be suitable for predicting bankruptcy. Thus, using the dynamic model, there is a possibility that bankruptcy prediction model is improved. In this paper, we propose RNN (Recurrent Neural Network) which is one of the deep learning methodologies. The RNN learns time series data and the performance is known to be good. Prior to experiment, we selected non-financial firms listed on the KOSPI, KOSDAQ and KONEX markets from 2010 to 2016 for the estimation of the bankruptcy prediction model and the comparison of forecasting performance. In order to prevent a mistake of predicting bankruptcy by using the financial information already reflected in the deterioration of the financial condition of the company, the financial information was collected with a lag of two years, and the default period was defined from January to December of the year. Then we defined the bankruptcy. The bankruptcy we defined is the abolition of the listing due to sluggish earnings. We confirmed abolition of the list at KIND that is corporate stock information website. Then we selected variables at previous papers. The first set of variables are Z-score variables. These variables have become traditional variables in predicting bankruptcy. The second set of variables are dynamic variable set. Finally we selected 240 normal companies and 226 bankrupt companies at the first variable set. Likewise, we selected 229 normal companies and 226 bankrupt companies at the second variable set. We created a model that reflects dynamic changes in time-series financial data and by comparing the suggested model with the analysis of existing bankruptcy predictive models, we found that the suggested model could help to improve the accuracy of bankruptcy predictions. We used financial data in KIS Value (Financial database) and selected Multivariate Discriminant Analysis (MDA), Generalized Linear Model called logistic regression (GLM), Support Vector Machine (SVM), Artificial Neural Network (ANN) model as benchmark. The result of the experiment proved that RNN's performance was better than comparative model. The accuracy of RNN was high in both sets of variables and the Area Under the Curve (AUC) value was also high. Also when we saw the hit-ratio table, the ratio of RNNs that predicted a poor company to be bankrupt was higher than that of other comparative models. However the limitation of this paper is that an overfitting problem occurs during RNN learning. But we expect to be able to solve the overfitting problem by selecting more learning data and appropriate variables. From these result, it is expected that this research will contribute to the development of a bankruptcy prediction by proposing a new dynamic model.