• Korea Trade Review
• /
• v.46 no.1
• /
• pp.169-187
• /
• 2021

This paper analyzes the feasibility of using machine learning and deep learning methods to forecast the income and employment using the strategic industries as well as investment, export, and exchange rates. The decision tree, artificial neural network, support vector machine, and deep learning models were used to forecast the income and employment in Busan. The following were the main findings of the comparison of their predictive abilities. First, the decision tree models predict the income and employment well. The forecasting values for the income and employment appeared somewhat differently according to the depth of decision trees and several conditions of strategic industries as well as investment, export, and exchange rates. Second, since the artificial neural network models show that the coefficients are somewhat low and RMSE are somewhat high, these models are not good forecasting the income and employment. Third, the support vector machine models show the high predictive power with the high coefficients of determination and low RMSE. Fourth, the deep neural network models show the higher predictive power with appropriate epochs and batch sizes. Thus, since the machine learning and deep learning models can predict the employment well, we need to adopt the machine learning and deep learning models to forecast the income and employment.

• Journal of the Korean Geographical Society
• /
• v.51 no.2
• /
• pp.219-233
• /
• 2016

In social science fields, statistical models are used almost exclusively for causal explanation, and explanatory modeling has been a mainstream until now. In contrast, predictive modeling has been rare in the fields. Hence, we focus on constructing the predictive non-parametric model, instead of the explanatory model. Gangnam-gu, Seoul was chosen as a study area and we collected single-family house sales data sold between 2011 and 2014. We applied non-parametric models proposed in machine learning area including generalized additive model(GAM), random forest, multivariate adaptive regression splines(MARS) and support vector machines(SVM). Models developed recently such as MARS and SVM were found to be superior in predictive power for house price estimation. Finally, spatial autocorrelation was accounted for in the non-parametric models additionally, and the result showed that their predictive power was enhanced further. We hope that this study will prompt methodology for property price estimation to be extended from traditional parametric models into non-parametric ones.

• International Journal of Control, Automation, and Systems
• /
• v.4 no.6
• /
• pp.669-681
• /
• 2006

In this work, a new method of design adaptive controllers for SISO systems based on multiple models and switching is presented. The controller selects the model from a given set, according to a switching rule based on output prediction errors. The goal is to design, at each sample instant, a predictive control law that ensures the robust stability of the closed-loop system and achieves the best performance for the current operating point. At each sample the proposed control scheme identifies a set of linear models that best characterizes the dynamics of the current operating region. Then, it carries out an automatic reconfiguration of the controller to achieve the best possible performance whilst providing a guarantee of robust closed-loop stability. The results are illustrated by simulations a nonlinear continuous and stirred tank reactor.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.44 no.4
• /
• pp.227-233
• /
• 2021

Predictive maintenance has been one of important applications of data science technology that creates a predictive model by collecting numerous data related to management targeted equipment. It does not predict equipment failure with just one or two signs, but quantifies and models numerous symptoms and historical data of actual failure. Statistical methods were used a lot in the past as this predictive maintenance method, but recently, many machine learning-based methods have been proposed. Such proposed machine learning-based methods are preferable in that they show more accurate prediction performance. However, with the exception of some learning models such as decision tree-based models, it is very difficult to explicitly know the structure of learning models (Black-Box Model) and to explain to what extent certain attributes (features or variables) of the learning model affected the prediction results. To overcome this problem, a recently proposed study is an explainable artificial intelligence (AI). It is a methodology that makes it easy for users to understand and trust the results of machine learning-based learning models. In this paper, we propose an explainable AI method to further enhance the explanatory power of the existing learning model by targeting the previously proposedpredictive model [5] that learned data from a core facility (Hyper Compressor) of a domestic chemical plant that produces polyethylene. The ensemble prediction model, which is a black box model, wasconverted to a white box model using the Explainable AI. The proposed methodology explains the direction of control for the major features in the failure prediction results through the Explainable AI. Through this methodology, it is possible to flexibly replace the timing of maintenance of the machine and supply and demand of parts, and to improve the efficiency of the facility operation through proper pre-control.

• 제어로봇시스템학회:학술대회논문집
• /
• 1993.10a
• /
• pp.754-759
• /
• 1993

An adaptive model predictive control (AMPC) strategy using auto-regression moving-average (ARMA) models is presented. The characteristic features of this methodology are the small computer memory requirement, high computational speed, robustness, and easy handling of nonlinear and time varying MIMO systems. Since the process dynamic behaviors are expressed by ARMA models, the model parameter adaptation is simple and fast to converge. The recursive least square (RLS) method with exponential forgetting is used to trace the process model parameters assuming the process is slowly time varying. The control performance of the AMPC is verified by both comparative simulation and experimental studies on distillation column control.

• The Journal of Asian Finance, Economics and Business
• /
• v.8 no.11
• /
• pp.1-10
• /
• 2021

This paper compares the performance of Fama-French three-factor and five-factor models using a dataset of 20 Pakistani commercial banks for the period 2011 to 2020. We focus on an emerging economy as the findings from earlier studies on developed countries cannot be generalized in emerging markets. For empirical analysis, twelve portfolios were developed based on size, market capitalization, investment strategy, and growth. Subsequently, we constructed five Fama-French factors namely, R_M, SMB, HML, RMW, and CMA. The OLS regression technique with robust standard errors was applied to compare the predictive power of both the Fama-French models. Further, we also compared the mean-variance efficiency of the Fama-French models through the GRS test. Our empirical analysis provides three unique and interesting findings. First, both asset pricing models have similar predictive power to explain the expected portfolio returns in most cases. Second, our results from the GRS test suggest that there is no noticeable difference in the mean-variance efficiency of one asset pricing model over the other. Third, we find that all factors of both Fama-French models are statistically significant and are important for explaining the volatility of expected commercial bank returns in the context of Pakistan.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.53 no.5
• /
• pp.699-706
• /
• 2020

Predictive models in food microbiology are used for predicting microbial growth or death rates using mathematical and statistical tools considering the intrinsic and extrinsic factors of food. This study developed predictive growth models for Bacillus cereus on dried laver Pyropia pseudolinearis stored at different temperatures (5, 10, 15, 20, and 25℃). Primary models developed for specific growth rate (SGR), lag time (LT), and maximum population density (MPD) indicated a good fit (R²≥0.98) with the Gompertz equation. The SGR values were 0.03, 0.08, and 0.12, and the LT values were 12.64, 4.01, and 2.17 h, at the storage temperatures of 15, 20, and 25℃, respectively. Secondary models for the same parameters were determined via nonlinear regression as follows: SGR=0.0228-0.0069*T₁+0.0005*T₁²; LT=113.0685-9.6256*T₁+0.2079*T₁²; MPD=1.6630+0.4284*T₁-0.0080*T₁² (where T₁ is the storage temperature). The appropriateness of the secondary models was validated using statistical indices, such as mean squared error (MSE<0.01), bias factor (0.99≤B_f≤1.07), and accuracy factor (1.01≤A_f≤1.14). External validation was performed at three random temperatures, and the results were consistent with each other. Thus, these models may be useful for predicting the growth of B. cereus on dried laver.

• Journal of Information Processing Systems
• /
• v.17 no.3
• /
• pp.630-644
• /
• 2021

Educational data analysis is attracting increasing attention with the rise of the big data industry. The amounts and types of learning data available are increasing steadily, and the information technology required to analyze these data continues to develop. The early identification of potential dropout students is very important; education is important in terms of social movement and social achievement. Here, we analyze educational data and generate predictive models for student dropout using logistic regression, a decision tree, a naïve Bayes method, and a multilayer perceptron. The multilayer perceptron model using independent variables selected via the variance analysis showed better performance than the other models. In addition, we experimentally found that not only grades but also extracurricular activities were important in terms of preventing student dropout.

• Communications for Statistical Applications and Methods
• /
• v.30 no.5
• /
• pp.437-451
• /
• 2023

Generalized linear models and generalized linear mixed models (GLMMs) are fundamental tools for predictive analyses. In insurance, GLMMs are particularly important, because they provide not only a tool for prediction but also a theoretical justification for setting premiums. Although thousands of resources are available for introducing GLMMs as a classical and fundamental tool in statistical analysis, few resources seem to be available for the insurance industry. This study targets insurance professionals already familiar with basic actuarial mathematics and explains GLMMs and their linkage with classical actuarial pricing tools, such as the Buhlmann premium method. Focus of the study is mainly on the modeling aspect of GLMMs and their application to pricing, while avoiding technical issues related to statistical estimation, which can be automatically handled by most statistical software.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.16 no.4
• /
• pp.416-422
• /
• 2006

Most of the existing statistical models of software manpower profile are based on the assumptions of the usage and development process. Therefore, there is no universally applicable estimation and prediction model. To develop a prediction model, this paper suggests the predictive filter as a prediction model for software manpower profile. Firs of all, we investigate the software manpower profile and we suggest the input-output of predictive filter and method for parameter determination. Then, its usefulness is empirically verified by analyzing the actual data obtained from the software projects. Based on the average relative prediction error and Pred(0.25), the suggested predictive filter is compared with other well-known statistical estimation models. As a result, the predictive filter generally has a simple structure and on the other hand, it adapts the software manpower profile very well.