• Proceedings of the Korean Institute of Building Construction Conference
• /
• 2007.11a
• /
• pp.109-113
• /
• 2007

This paper introduces a system, Simulation based Stochastic Markup Estimation System (S2ME), for estimating optimum markup for a project. The system was designed and implemented to better represent the real world system involved in construction bidding. The findings obtained from the analysis of existing assumptions used in the previous quantitative markup estimation methods were incorporated to improve the accuracy and predictability of the S2ME. The existing methods has four categories of assumption as follows; (1) The number of competitors and who is the competitors are known, (2) A typical competitor, who is fictitious, is assumed for easy computation, (3) the ratio of bid price against cost estimate (B/C) is assumed to follow normal distribution, (4) The deterministic output obtained from the probabilistic equation of existing models is assumed to be acceptable. However, these assumptions compromise the accuracy of prediction. In practice, the bidding patterns of the bidders are randomized in competitive bidding. To complement the lack of accuracy contributed by these assumptions, bidding project was randomly selected from the pool of bidding database in the simulation experiment. The probability to win the bid in the competitive bidding was computed using the profile of the competitors appeared in the selected bidding project record. The expected profit and probability to win the bid was calculated by selecting a bidding record randomly in an iteration of the simulation experiment under the assumption that the bidding pattern retained in historical bidding DB manifest revival. The existing computation, which is handled by means of deterministic procedure, were converted into stochastic model using simulation modeling and analysis technique as follows; (1) estimating the probability distribution functions of competitors' B/C which were obtained from historical bidding DB, (2) analyzing the sensitivity against the increment of markup using normal distribution and actual probability distribution estimated by distribution fitting, (3) estimating the maximum expected profit and optimum markup range. In the case study, the best fitted probability distribution function was estimated using the historical bidding DB retaining the competitors' bidding behavior so that the reliability was improved by estimating the output obtained from simulation experiment.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.15 no.2
• /
• pp.43-51
• /
• 2011

Probabilistic seismic hazard analysis (PSHA), which can effectively apply inevitable uncertainties in seismic data, considers a number of seismotectonic models and attenuation equations. The calculated hazard by PSHA is generally a value dependent on peak ground acceleration (PGA) and expresses the value as an annual exceedance probability. To represent the uncertainty range of a hazard which has occurred using various seismic data, a hazard curve figure shows both a mean curve and percentile curves (15, 50, and 85). The percentile performs an important role in that it indicates the uncertainty range of the calculated hazard, could be calculated using various methods by the relation of the weight and hazard. This study using the weight accumulation method, the weighted hazard method, the maximum likelihood method, and the moment method, has calculated the percentile of the computed hazard by PSHA on the Shinuljin 1, 2 site. The calculated percentile using the weight accumulation method, the weighted hazard method, and the maximum likelihood method, have similar trends and represent the range of all computed hazards by PSHA. The calculated percentile using the moment method effectively showed the range of hazards at the source which includes a site. This study suggests the moment method as effective percentile calculation method considering the almost same mean hazard for the seismotectonic model and a source which includes a site.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.23 no.3
• /
• pp.258-264
• /
• 2011

A MCS technique is represented to stochastically analyze the uncertainties of wave forces exerted on the upright sections of composite breakwaters. A stochastical models for horizontal and uplift wave forces can be straightforwardly formulated as a function of the probabilistic characteristics of maximum wave height. Under the assumption of wave forces followed by extreme distribution, the behaviors of relative wave forces to Goda's wave forces are studied by the MCS technique. Double-truncated normal distribution is applied to take the effects of uncertainties of scale and shape parameters of extreme distribution into account properly. Averages and variances of relative wave forces are quantitatively calculated with respect to the exceedance probabilities of maximum design wave height. It is found that the averages of relative wave forces may be decreased consistently with the increases of the exceedance probabilities. In particular, the averages on uplift wave force are evaluated slightly larger than those on horizontal wave force, but the variations of coefficient of the former are adversely smaller than those of the latter. It means that the uncertainties of uplift wave forces are smaller than those of horizontal wave forces in the same condition of the exceedance probabilities. Therefore, the present results could be useful to the reliability based-design method that require the statistical properties about the uncertainties of wave forces.

• Journal of Korea Water Resources Association
• /
• v.39 no.3 s.164
• /
• pp.275-288
• /
• 2006

This study develops ESP (Ensemble Streamflow Prediction) system by using medium-term numerical weather prediction model which is GDAPS(T213) of KMA. The developed system forecasts medium- and long-range exceedance Probability for streamflow and RPSS evaluation scheme is used to analyze the accuracy of probability forecasts. It can be seen that the daily probability forecast results contain high uncertainties. A sensitivity analysis with respect to forecast time resolution shows that uncertainties decrease and accuracy generally improves as the forecast time step increase. Weekly ESP results by using the GDAPS output with a lead time of up to 28 days are more accurately predicted than traditional ESP results because conditional probabilities are stably distributed and uncertainties can be reduced. Therefore, it can be concluded that the developed system will be useful tool for medium- and long-term reservoir inflow forecasts in order to manage water resources.

• Journal of KIISE:Databases
• /
• v.28 no.4
• /
• pp.688-701
• /
• 2001

Web caching has become an important problem addressing the performance issues in web applications. In this paper we propose a method that enhances the performance of web caching by incorporating the expiration time of web data we introduce the notion of the effective reference probability that incorporates the effect of expiration time into the reference probability used in the existing cache replacement algorithms .We formally define the effective reference probability and derive it theoretically using a probabilistic model. By simply replacing probabilities with the effective reference probability in the existing cache replacement algorithms we can take the effect of expiration time into account The results of performance evaluation through experiments show that the replacement algorithms using the effective reference probability always outperform the existing ones. The reason is that the proposed method precisely reflects the theoretical probability of getting the cache effect, and thus, incorporates the influence of the expiration time more effectively. In particular when the cache fraction is 0.05 and data update is comparatively frequent (i.e. the update frequency is more than 1/0 of the reference frequency) the performance enhancement is more than 30% in LRU-2 and 13% in Aggarwal's method (PSS integrating a refresh overhead factor) The results show that effective reference probability contributes significantly to the performance enhancement of the web cache in the presence of expiration time.

• Journal of Korea Water Resources Association
• /
• v.54 no.10
• /
• pp.769-777
• /
• 2021

This study investigated the change of the achievement rate of the target water quality conditioned on the occurrence of severe drought, to assess the effects of meteorological drought on the water quality management in the Nakdong River basin. Using three drought indices with difference time scales such as 30-, 60-, 90-day, i.e., SPI30, SPI60, SPI90, and three water quality indicators such as biochemical oxygen demand (BOD), total organic carbon (TOC), and total phosphorus (T-P), we first analyzed the relationship between severe drought occurrence water quality change in mid-sized watersheds, and identified the watersheds in which water quality was highly affected by severe drought. The Bayesian network models were constructed for the watersheds to probabilistically assess the relationship between severe drought and water quality management. Among 22 mid-sized watersheds in the Nakdong River basin, four watersheds, such as #2005, #2018, #2021, and #2022, had high environmental vulnerability to severe drought. In addition, severe drought affected spring and fall water quality in the watershed #2021, summer water quality in the #2005, and winter water quality in the #2022. The causal relationship between drought and water quality management is usufaul in proactive drought management.

• Korean Journal of Remote Sensing
• /
• v.37 no.6_2
• /
• pp.1869-1880
• /
• 2021

Korean Journal of Remote Sensing (KJRS), leading the field of remote sensing and GIS in South Korea for over 37 years, has published interdisciplinary research papers. In this study, we performed the topic modeling based on Latent Dirichlet Allocation (LDA), a probabilistic generative model, to identify the research topics and trends using 1) the whole articles, and 2) specific articles related to natural and environmental disasters published in KJRS by analyzing titles, keywords, and abstracts. The results of LDA showed that 4 topics('Polar', 'Hydrosphere', 'Geosphere', and 'Atmosphere') were identified in the whole articles and the topic of 'Polar' was dominant among them (linear slope=3.51 × 10^-3, p<0.05) over time. For the specific articles related to natural and environmental disasters, the optimal number of topics were 7 ('Marine pollution', 'Air pollution', 'Volcano', 'Wildfire', 'Flood', 'Drought', and 'Heavy rain') and the topic of 'Air pollution' was dominant (linear slope=2.61 × 10^-3, p<0.05) over time. The results from this study provide the history and insight into natural and environmental disasters in KRJS with multidisciplinary researchers.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.12
• /
• pp.87-96
• /
• 2020

To use a hydraulic structure effectively, the velocity of a river should be known in detail. In reality, velocity measurements are not conducted sufficiently because of their high cost. The formulae to yield the flux and velocity of the river are commonly called the Manning and Chezy formulae, which are empirical equations applied to uniform flow. This study is based on Chiu (1987)'s paper using entropy theory to solve the limits of the existing velocity formula and distribution and suggests the velocity and distance formula derived from information entropy. The data of a channel having records of a spot's velocity was used to verify the derived formula's utility and showed R2 values of distance and velocity of 0.9993 and 0.8051~0.9483, respectively. The travel distance and velocity of a moving spot following the streamflow were calculated using some flow information, which solves the difficulty in frequent flood measurements when it is needed. This can be used to make a longitudinal section of a river composed of a horizontal distance and elevation. Moreover, GIS makes it possible to obtain accurate information, such as the characteristics of a river. The connection with flow information and GIS model can be used as alarming and expecting flood systems.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.23 no.4
• /
• pp.391-404
• /
• 2021

This study developed a long-term prediction model for the potential yield of garlic based on a genetic algorithm (GA) by utilizing global climate reanalysis data. The GA is used for digging the inherent signals from global climate reanalysis data which are both directly and indirectly connected with the garlic yield potential. Our results indicate that both deterministic and probabilistic forecasts reasonably capture the inter-annual variability of crop yields with temporal correlation coefficients significant at 99% confidence level and superior categorical forecast skill with a hit rate of 93.3% for 2 × 2 and 73.3% for 3 × 3 contingency tables. Furthermore, the GA method, which considers linear and non-linear relationships between predictors and predictands, shows superiority of forecast skill in terms of both stability and skill scores compared with linear method. Since our result can predict the potential yield before the start of farming, it is expected to help establish a long-term plan to stabilize the demand and price of agricultural products and prepare countermeasures for possible problems in advance.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.3
• /
• pp.133-140
• /
• 2022

We investigate the performance of deep learning-based Korean language models on a task of predicting the score range of Korean essays written by foreign students. We construct a data set containing a total of 304 essays, which include essays discussing the criteria for choosing a job ('job'), conditions of a happy life ('happ'), relationship between money and happiness ('econ'), and definition of success ('succ'). These essays were labeled according to four letter grades (A, B, C, and D), and a total of eleven essay score range prediction experiments were conducted (i.e., five for predicting the score range of 'job' essays, five for predicting the score range of 'happiness' essays, and one for predicting the score range of mixed topic essays). Three deep learning-based Korean language models, KoBERT, KcBERT, and KR-BERT, were fine-tuned using various training data. Moreover, two traditional probabilistic machine learning classifiers, naive Bayes and logistic regression, were also evaluated. Experiment results show that deep learning-based Korean language models performed better than the two traditional classifiers, with KR-BERT performing the best with 55.83% overall average prediction accuracy. A close second was KcBERT (55.77%) followed by KoBERT (54.91%). The performances of naive Bayes and logistic regression classifiers were 52.52% and 50.28% respectively. Due to the scarcity of training data and the imbalance in class distribution, the overall prediction performance was not high for all classifiers. Moreover, the classifiers' vocabulary did not explicitly capture the error features that were helpful in correctly grading the Korean essay. By overcoming these two limitations, we expect the score range prediction performance to improve.