• Korean Journal of Remote Sensing
• /
• v.36 no.4
• /
• pp.573-586
• /
• 2020

PM (particulate matter) is of interest to everyone because it can have adverse effects on human health by the infiltration from respiratory to internal organs. To date, many studies have made efforts for the prediction of PM10 and PM2.5 concentrations. Unlike previous studies, we conducted the prediction of tomorrow's PM10 concentration for the Air Korea stations using Chinese PM10 data in addition to the satellite AOD and weather variables. We constructed 230,639 matchups from the raw data over 3 million and built an RF (random forest) model from the matchups to cope with the complexity and nonlinearity. The validation statistics from the blind test showed excellent accuracy with the RMSE (root mean square error) of 9.905 ㎍/㎥ and the CC (correlation coefficient) of 0.918. Moreover, our prediction model showed a stable performance without the dependency on seasons or the degree of PM10 concentration. However, part of coastal areas had a relatively low accuracy, which implies that a dedicated model for coastal areas will be necessary. Additional input variables such as wind direction, precipitation, and air stability should also be incorporated into the prediction model as future work.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.12 no.1
• /
• pp.428-439
• /
• 2020

The continuous development of information and communication technologies has resulted in an exponential increase in data. Consequently, technologies related to data analysis are growing in importance. The shipbuilding industry has high production uncertainty and variability, which has created an urgent need for data analysis techniques, such as machine learning. In particular, the industry cannot effectively respond to changes in the production-related standard time information systems, such as the basic cycle time and lead time. Improvement measures are necessary to enable the industry to respond swiftly to changes in the production environment. In this study, the lead times for fabrication, assembly of ship block, spool fabrication and painting were predicted using machine learning technology to propose a new management method for the process lead time using a master data system for the time element in the production data. Data preprocessing was performed in various ways using R and Python, which are open source programming languages, and process variables were selected considering their relationships with the lead time through correlation analysis and analysis of variables. Various machine learning, deep learning, and ensemble learning algorithms were applied to create the lead time prediction models. In addition, the applicability of the proposed machine learning methodology to standard work hour prediction was verified by evaluating the prediction models using the evaluation criteria, such as the Mean Absolute Percentage Error (MAPE) and Root Mean Squared Logarithmic Error (RMSLE).

• Journal of the Korean Society for Precision Engineering
• /
• v.15 no.7
• /
• pp.129-138
• /
• 1998

The objective of this study is to demonstrate the ability of a computer simulation of microstructural evolution in hot forging of C-Mn steels. The development of microstructure is strongly dependent on process variables and metallurgical factors that affect time history of thermodynamical variables such as temperature, strain. and strain rate during deformation. Then finite element method is applied for the prediction of microstructural evolution, and it should be coupled with heat transfer analysis to consider the change of thermodynamical properties during forming process. In this study, Yada's recrystallization model and rigid-thermoviscoplastic finite element method are employed in order to analyze microstructural evolution during hot forging process. To show the validity and effectiveness of the proposed method, experiments are accomplished and the results of experiments are compared with those of simulations.

• Journal of the Korean Data and Information Science Society
• /
• v.26 no.5
• /
• pp.1087-1095
• /
• 2015

Influenza, commonly known as "the flu", is an infectious disease caused by the influenza virus. We consider, in this paper, regression models as a prediction model of influenza disease. While most of previous researches use mainly the meteorological variables as a predictive variables, we consider social media information in the models. As a result, we found that the contributions of two-type of informations are comparable. We used the medical treatment data of influenza provided by Natioal Health Insurance Survice (NHIS) and the meteorological data provided by Korea Meteorological Administration (KMA). We collect social media information (twitter buzz amount) from Twitter. Time series model is also considered for comparison.

• Proceedings of the Korean Institute of Building Construction Conference
• /
• 2020.11a
• /
• pp.69-70
• /
• 2020

Accidents in the construction industry are very high compared to other industries, and the number is also increasing steeply every year. Relevant studies were limited for solving the problems. The purpose of this study is to develop a comprehensive risk prediction process for personnel deployed at construction sites on safety management. First of all, the variables were divided into fixed, real-time and working types variables, and the relevant comprehensive data were collected. Second, the probability of a disaster was derived based on the collected data, and weights for each variable were calculated using the dummy regression analysis method using statistical methodology. Lastly, the resulting weighting and disaster probability equation was constructed, and The Final Risk Calculation Formula was developed. The Final Risk Calculation Formula presented in this study is expected to have a significant impact on the establishment of effective safety management measures to prevent possible safety accidents at construction sites

• Journal of the Korean Society of Safety
• /
• v.33 no.4
• /
• pp.98-104
• /
• 2018

Macroscopic accident analyses have been conducted to incorporate transportation safety into long-term transportation planning. In macro-level accident prediction model, exposure variable(e.g. a settled population) have been used as fundamental explanatory variable under the concept that each trip will be subjected to a probable risk of accident. However, a settled population may be embedded error by exclusion of active population concept. The objective of this research study is to develop macro-level accident prediction model using floating population variable(concept of including a settled population and active population) collected from mobile phone data. The concept of accident prediction models is introduced utilizing exposure variable as explanatory variable in a generalized linear regression with assumption of a negative binomial error structure. The goodness of fit of model using floating population variable is compared with that of the each models using population and the number of household variables. Also, log transformation models are additionally developed to improve the goodness of fit. The results show that the log transformation model using floating population variable is useful for capturing the relationships between accident and exposure variable and generally perform better than the models using other existing exposure variables. The developed model using floating population variable can be used to guide transportation safety policy decision makers to allocate resources more efficiently for the regions(or zones) with higher risk and improve urban transportation safety in transportation planning step.

• Journal of Welding and Joining
• /
• v.32 no.2
• /
• pp.43-47
• /
• 2014

Famous artificial neural network (ANN) is applied to predict proper process window of arc welding. Target weldment is variously combined lap joint fillet welding of automotive steel plates. ANN's system variable such as number of hidden layers, perceptrons and transfer function are carefully selected through case by case test. Input variables are welding condition and steel plate combination, for example, welding machine type, shield gas composition, current, speed and strength, thickness of base material. The number of each input variable referred in welding experiment is counted and provided to make it possible to presume the qualitative precision and limit of prediction. One of experimental process windows is excluded for predictability estimation and the rest are applied for neural network training. As expected from basic ANN theory, experimental condition composed of frequently referred input variables showed relatively more precise prediction while rarely referred set showed poorer result. As conclusion, application of ANN to arc welding process window derivation showed comparatively practical feasibility while it still needs more training for higher precision.

• Geomechanics and Engineering
• /
• v.31 no.6
• /
• pp.637-651
• /
• 2022

The seismic design of embankment dams requires more comprehensive studies to understand the behaviour of dams. Deformations primarily control this behaviour occur during or after earthquake loading. Dam failures and incidents show that the impacts of deformations should be reviewed for existing and new embankment dams. Overtopping erosion failure can occur if crest deformations exceed the freeboard at the time of the deformations. Therefore, crest settlement is one of the most critical deformations. This study developed empirical formulas using Gene Expression Programming (GEP) based on 88 cases. In the analyses, dam height (H_d), alluvium thickness (H_a), the magnitude-acceleration-factor (MAF) values developed based on earthquake magnitude (Mw) and peak ground acceleration (PGA) within this study have been chosen as variables. Results show that GEP models developed in the paper are remarkably robust and accessible tools to predict earthquake-induced crest settlement of embankment dams and perform superior to the existing formulation. Also, dam engineering professionals can use them practically because the variables of prediction equations are easily accessible after the earthquake.

• Journal of the Korean Geotechnical Society
• /
• v.33 no.6
• /
• pp.27-36
• /
• 2017

Gravel compaction pile method has been widely used to improve the soft ground on the land or sea as one of the soft ground improvement technique. The ultimate bearing capacity of the ground reinforced by gravel compaction piles is affected by the soil strength, the replacement ratio of pile, construction conditions, and so on, and various prediction equations have been proposed to predict this. However, the prediction of the ultimate bearing capacity using the existing models has a very large error and variation, and it is not suitable for practical design. In this study, multiple regression analysis was performed using field loading test results to predict the ultimate bearing capacity of ground reinforced by gravel compaction pile, and the most efficient input variables are selected through evaluation of error by leave one out cross validation, and a multiple regression equation for the prediction of ultimate bearing capacity was proposed. In addition, the prediction error was evaluated by applying artificial neural network using the selected input variables, and the results were compared with those of the existing model.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.14 no.5
• /
• pp.377-384
• /
• 2002

For economic evaluation of cooling plant equipments, it is necessary to simplify energy Prediction method, which should includes efficiency corrected by part-load ratio. This study proposed simplified method with regression equations of time-average partial loads and refrigerator capacity. DOE-2 Program was used to carry out a parametric study of twelve design variables. Five input variables were considered to be significant and were used in the regression equations. To test accuracy of simplified method, calculated results were compared with DOE-2 simulated results. Test result showes a good agreement with the simulation result with an error of 5.9∼7.6%. It is expected that this method can be used as an easy prediction tool for comparing energy use of different cooling plants during the early design stage.