• The Journal of Engineering Geology
• /
• v.29 no.4
• /
• pp.417-425
• /
• 2019

Agricultural groundwater use accounts for 51.8% of total groundwater use, so accurate estimation of groundwater use is important for efficient groundwater management. The purpose of this study is to develop a method for estimating the groundwater use of agricultural (rice farming and dry-field farming) wells using regression tree model based on the measured data of 370 wells. Three input variables of the model were evaluated as being significant: well depth, pipe diameter, and pump capacity, and the importance of each variable was 75% for well depth, 17% for pipe diameter, and 8% for pumping capacity. The daily usage of agricultural (rice farming and dry-field farming) wells by the regression tree model was estimated to be very similar to the actual usage, compared to the previous estimation method proposed by the Ministry of Construction and Transportation. In the future, it is expected that the reliability of the usage statistics will be improved if additional observed data is secured and this classification method is modified.

• Journal of Korea Society of Industrial Information Systems
• /
• v.14 no.3
• /
• pp.106-113
• /
• 2009

Various statistical classification methods have been used to establish prediction model of underground cultural assets in our country. Among them, linear discriminant analysis, logistic regression, decision tree, neural network, and support vector machines are used in this paper. We introduced the basic concepts of above-mentioned classification methods and applied these to the analyses of real data of I city. As a results, five different prediction models are suggested. And also model comparisons are executed by suggesting correct classification rates of the fitted models. To see the applicability of the suggested models for a new data set, simulations are carried out. R packages and programs are used in real data analyses and simulations. Especially, the detailed executing processes by R are provided for the other analyser of related area.

• Annual Conference on Human and Language Technology
• /
• 1999.10e
• /
• pp.191-195
• /
• 1999

자연스럽고 명료한 한국어 Text-to-Speech 변환 시스템을 위해서 음소의 지속 시간을 제어하는 일은 매우 중요하다. 음소의 지속 시간은 여러 가지 문맥 정보에 의해서 변화하므로 제어 규칙에 의존하기 보다 방대한 데이터베이스를 이용하여 통계적인 기법으로 음소의 지속 시간에 변화를 주는 요인을 찾아내려고 하는 것이 지금의 추세이다. 본 연구에서도 트리기반 모델링 방법중의 하나인 CART(classification and regression tree) 방법을 사용하여 회귀 트리를 생성하고, 생성된 트리에 기반하여 음소의 지속 시간 예측 모델과, 자연스러운 끊어 읽기를 위한 휴지 기간 예측 모델을 제안하고 있다. 실험에 사용한 음성코퍼스는 550개의 문장으로 구성되어 있으며, 이 중 428개 문장으로 회귀 트리를 학습시켰고, 나머지 122개의 문장으로 실험하였다. 모델의 평가를 위해서 실제값과 예측값과의 상관관계를 구하였더니 음소의 지속 시간을 예측하는 회귀 트리에서는 상관계수가 0.84로 계산되었고, 끊어 읽는 경계에서의 휴지 기간을 예측하는 회귀 트리에서는 상관계수가 0.63으로 나타났다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.8
• /
• pp.1560-1566
• /
• 2017

Recently, accumulation of data on pig farm is enabled through the wide spread of smart pig farm equipped with Internet-of-Things based sensors, and various machine learning algorithms are applied on the data in order to improve the productivity of pig farm. Herein, multiple machine learning schemes are used to predict the water usage in pig farm which is known to be one of the most important element in pig farm management. Especially, regression algorithms, which are linear regression, regression tree and AdaBoost regression, and classification algorithms which are logistic classification, decision tree and support vector machine, are applied to derive a prediction scheme which forecast the water usage based on the temperature and humidity of pig farm. Through performance evaluation, we find that the water usage can be predicted with high accuracy. The proposed scheme can be used to detect the malfunction of water system which prevents the death of pigs and reduces the loss of pig farm.

• Tunnel and Underground Space
• /
• v.31 no.6
• /
• pp.508-519
• /
• 2021

In this study, a model was developed to predict the peak particle velocity (PPV) that affects people and the surrounding environment during blasting. Four machine learning models using the k-nearest neighbors (kNN), classification and regression tree (CART), support vector regression (SVR), and particle swarm optimization (PSO)-SVR algorithms were developed and compared with each other to predict the PPV. Mt. Yogmang located in Changwon-si, Gyeongsangnam-do was selected as a study area, and 1048 blasting data were acquired to train the machine learning models. The blasting data consisted of hole length, burden, spacing, maximum charge per delay, powder factor, number of holes, ratio of emulsion, monitoring distance and PPV. To evaluate the performance of the trained models, the mean absolute error (MAE), mean square error (MSE), and root mean square error (RMSE) were used. The PSO-SVR model showed superior performance with MAE, MSE and RMSE of 0.0348, 0.0021 and 0.0458, respectively. Finally, a method was proposed to predict the degree of influence on the surrounding environment using the developed machine learning models.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2007.11a
• /
• pp.886-890
• /
• 2007

Machine fault prognosis techniques have been considered profoundly in the recent time due to their profit for reducing unexpected faults or unscheduled maintenance. With those techniques, the working conditions of components, the trending of fault propagation, and the time-to-failure are forecasted precisely before they reach the failure thresholds. In this work, we propose an approach of Least Square Regression Tree (LSRT), which is an extension of the Classification and Regression Tree (CART), in association with one-step-ahead prediction of time-series forecasting technique to predict the future conditions of machines. In this technique, the number of available observations is firstly determined by using Cao's method and LSRT is employed as prognosis system in the next step. The proposed approach is evaluated by real data of low methane compressor. Furthermore, the comparison between the predicted results of CART and LSRT are carried out to prove the accuracy. The predicted results show that LSRT offers a potential for machine condition prognosis.

• The Korean Journal of Applied Statistics
• /
• v.24 no.5
• /
• pp.847-859
• /
• 2011

cDNA microarray-based comparative genomic hybridization(CGH) data includes low-intensity spots and thus a statistical strategy is needed to detect subtle differences between different cancer classes. In this study, genes displaying a high frequency of alteration in one of the different classes were selected among the pre-selected genes that show relatively large variations between genes compared to total variations. Utilizing copy-number changes of the selected genes, this study suggests a statistical approach to predict patients' classes with increased performance by pre-classifying patients with similar genetic alteration scores. Two-stage logistic regression model(TLRM) was suggested to pre-classify homogeneous patients and predict patients' classes for cancer prediction; a decision tree(DT) was combined with logistic regression on the set of informative genes. TLRM was constructed in cDNA microarray-based CGH data from the Cancer Metastasis Research Center(CMRC) at Yonsei University; it predicted the patients' clinical diagnoses with perfect matches (except for one patient among the high-risk and low-risk classified patients where the performance of predictions is critical due to the high sensitivity and specificity requirements for clinical treatments. Accuracy validated by leave-one-out cross-validation(LOOCV) was 83.3% while other classification methods of CART and DT performed as comparisons showed worse performances than TLRM.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.34 no.3
• /
• pp.29-41
• /
• 2009

The purpose of this paper is to develop a new classification method for predicting consumer choice based on genetic algorithm, and to validate Its prediction power over existing methods. To serve this purpose, we propose a hybrid model, and discuss Its methodological characteristics in comparison with other existing classification methods. Also, we conduct a series of experiments employing survey data of consumer choices of MP3 players to assess the prediction power of the model. The results show that the suggested model in this paper is statistically superior to the existing methods such as logistic regression model, artificial neural network model and decision tree model in terms of prediction accuracy. The model is also shown to have an advantage of providing several strategic information of practical use for consumer choice.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2008.10a
• /
• pp.1-7
• /
• 2008

The purpose of this paper is to develop a new classification method for predicting consumer choice based on genetic algorithm, and to validate its prediction power over existing methods. To serve this purpose, we propose a hybrid model, and discuss its methodological characteristics in comparison with other existing classification methods. Also, to assess the prediction power of the model, we conduct a series of experiments employing survey data of consumer choices of MP3 players. The results show that the suggested model in this paper is statistically superior to the existing methods such as logistic regression model, artificial neural network model and decision tree model in terms of prediction accuracy. The model is also shown to have an advantage of providing several strategic information of practical use for consumer choice.

• Journal of the Korean Data and Information Science Society
• /
• v.23 no.2
• /
• pp.375-383
• /
• 2012

The purpose of the ensemble methods is to increase the accuracy of prediction through combining many classifiers. According to recent studies, it is proved that random forests and forward stagewise regression have good accuracies in classification problems. However they have great prediction error in separation boundary points because they used decision tree as a base learner. In this study, we use the kernel ridge regression instead of the decision trees in random forests and boosting. The usefulness of our proposed ensemble methods was shown by the simulation results of the prostate cancer and the Boston housing data.