• Journal of the Korean Data and Information Science Society
• /
• 제23권2호
• /
• 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.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2010년도 추계학술발표대회
• /
• pp.596-598
• /
• 2010

본 논문에서는 의료영상 중 X-ray 영상을 대상으로 영상을 분류하고 분류 결과에 따라 다중 키워드를 생성하는 방법을 제시한다. X-ray영상은 대부분 그레이 영상임으로 Local Binary Patterns (LBP)을 이용하여 픽셀간의 연관성을 특징으로 추출하고, 실시간 학습 및 분류가 가능한 Random Forests 분류기로 영상들을 30개의 클래스로 분류한다. 또한, 미리 정의된 신체 부위간의 관계 가중치를 분류 스코어에 결합하여 신뢰값을 생성하고 이를 기반으로 영상에 대해 다중 주석을 부여하게 된다. 이렇게 부여된 다중 주석은 키워드 기반의 의료영상을 가능케 함으로 보다 쉽고 효율적인 검색 환경을 제공할 수 있다.

• 응용통계연구
• /
• 제22권4호
• /
• pp.771-780
• /
• 2009

백내장 질환은 노령인구가 증가하고 있는 시점에서 사회, 경제적으로 심각한 문제로 부각되고 있는 질병으로 조기 진단이 이루어진다면 발병률을 크게 줄일 수 있는 질병이다. 본 연구에서는 백내장을 조기 진단하기 위한 예측 모형을 구축하고자 1994년부터 2001년까지 연세대학병원에서 2회 이상 건강검진을 받고 의사진단을 통해 백내장 여부를 확인할 수 있는 30세 이상 남 녀 3,237명에 대한 건강검진 수검 자료를 활용하여 백내장 발생 위험 예측모형을 개발하였다. 모형개발에는 데이터마이닝 기법인 Random Forests를 사용하였고, 기존의 로지스틱 회귀분석, 판별분석, 의사결정나무 모형(Decision tree), 나이브베이즈(Naive Bayes), 앙상블 모형인 배깅(Bagging)과 아킹(Arcing)을 이용하여 그 성능을 비교 분석하였다. Random Forests를 통해 개발한 백내장 발생 예측모형은 정확도가 67.16%, 민감도가 72.28%였고, 주요 영향요인은 연령, 혈당, 백혈구수치(WBC), 혈소판수치(platelet), 중성지질(triglyceride), BMI였다. 이 결과는 의사의 안과검진 정보 없이 건강검진 수검 자료만으로 백내장 질환 유 무에 관한 정보를 70% 정도 예측할 수 있음을 보여주는 것으로, 백내장의 조기 진단에 많은 기여를 할 것으로 판단된다.

• Management Science and Financial Engineering
• /
• 제20권1호
• /
• pp.1-9
• /
• 2014

In multi-objective scheduling problems, the objectives are usually in conflict. To obtain a satisfactory compromise and resolve the issue of NP-hardness, most existing works have suggested employing meta-heuristic methods, such as genetic algorithms. In this research, we propose a novel data-driven approach for generating a single solution that compromises multiple rules pursuing different objectives. The proposed method uses a data mining technique, namely, random forests, in order to extract the logics of several historic schedules and aggregate those. Since it involves learning predictive models, future schedules with the same previous objectives can be easily and quickly obtained by applying new production data into the models. The proposed approach is illustrated with a simulation study, where it appears to successfully produce a new solution showing balanced scheduling performances.

• Asia pacific journal of information systems
• /
• 제26권1호
• /
• pp.66-79
• /
• 2016

The availability of detailed data on customers' online behaviors and advances in big data analysis techniques enable us to predict consumer behaviors. In the past, researchers have built purchase prediction models by analyzing clickstream data; however, these clickstream-based prediction models have had several limitations. In this study, we propose a new method for purchase prediction that combines information theory with machine learning techniques. Clickstreams from 5,000 panel members and data on their purchases of electronics, fashion, and cosmetics products were analyzed. Clickstreams were summarized using the 'entropy' concept from information theory, while 'random forests' method was applied to build prediction models. The results show that prediction accuracy of this new method ranges from 0.56 to 0.83, which is a significant improvement over values for clickstream-based prediction models presented in the past. The results indicate further that consumers' information search behaviors differ significantly across product categories.

• 응용통계연구
• /
• 제24권4호
• /
• pp.587-595
• /
• 2011

Credit scoring is an objective and automatic system to assess the credit risk of each customer. The logistic regression model is one of the popular methods of credit scoring to predict the default probability; however, it may not detect possible nonlinear features of predictors despite the advantages of interpretability and low computation cost. In this paper, we propose to use a generalized partially linear model as an alternative to logistic regression. We also introduce modern ensemble technologies such as bagging, boosting and random forests. We compare these methods via a simulation study and illustrate them through a German credit dataset.

• Communications for Statistical Applications and Methods
• /
• 제16권1호
• /
• pp.185-193
• /
• 2009

Machine learning methods such as support vector machines and random forests yield nonparametric prediction functions of the form y = $f(x_1,{\ldots},x_p)$. As a sequel to the previous article (Huh and Lee, 2008) for visualizing nonparametric functions, I propose more sensible graphs for visualizing y = $f(x_1,{\ldots},x_p)$ herein which has two clear advantages over the previous simple graphs. New graphs will show a small number of prototype curves of $f(x_1,{\ldots},x_{j-1},x_j,x_{j+1}{\ldots},x_p)$, revealing statistically plausible portion over the interval of $x_j$ which changes with ($x_1,{\ldots},x_{j-1},x_{j+1},{\ldots},x_p$). To complement the visual display, matching importance measures for each of p predictor variables are produced. The proposed graphs and importance measures are validated in simulated settings and demonstrated for an environmental study.

• Communications for Statistical Applications and Methods
• /
• 제24권6호
• /
• pp.543-559
• /
• 2017

Tree-based regression and classification ensembles form a standard part of the data-science toolkit. Many commonly used methods take an algorithmic view, proposing greedy methods for constructing decision trees; examples include the classification and regression trees algorithm, boosted decision trees, and random forests. Recent history has seen a surge of interest in Bayesian techniques for constructing decision tree ensembles, with these methods frequently outperforming their algorithmic counterparts. The goal of this article is to survey the landscape surrounding Bayesian decision tree methods, and to discuss recent modeling and computational developments. We provide connections between Bayesian tree-based methods and existing machine learning techniques, and outline several recent theoretical developments establishing frequentist consistency and rates of convergence for the posterior distribution. The methodology we present is applicable for a wide variety of statistical tasks including regression, classification, modeling of count data, and many others. We illustrate the methodology on both simulated and real datasets.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권3호
• /
• pp.1566-1582
• /
• 2019

Silent Data Corruptions (SDCs) is a serious reliability issue in many domains of computer system. The identification and protection of the program instructions that cause SDCs is one of the research hotspots in computer reliability field at present. A lot of solutions have already been proposed to solve this problem. However, many of them are hard to be applied widely due to time-consuming and expensive costs. This paper proposes an intelligent approach named SDCPredictor to identify the instructions that cause SDCs. SDCPredictor identifies SDC-causing Instructions depending on analyzing the static and dynamic features of instructions rather than fault injections. The experimental results demonstrate that SDCPredictor is highly accurate in predicting the SDCs proneness. It can achieve higher fault coverage than previous similar techniques in a moderate time cost.

• Journal of Forest and Environmental Science
• /
• 제36권2호
• /
• pp.113-121
• /
• 2020

This study was carried out in degraded and non-degraded community forests (CF) in the Terai region of Kanchanpur district, Nepal. A total of 63 concentric sample plots each of 500 ㎡ was laid in the inventory for estimating above and below-ground biomass of forests by using systematic random sampling with a sampling intensity of 0.5%. Mallotus philippinensis and Shorea robusta were the most dominant species in degraded and non-degraded CF accounting Importance Value Index (I.V.I) of 97.16 and 178.49, respectively. Above-ground tree biomass carbon in degraded and non-degraded community forests was 74.64±16.34 t ha^-1 and 163.12±20.23 t ha^-1, respectively. Soil carbon sequestration in degraded and non-degraded community forests was 42.55±3.10 t ha^-1 and 54.21±3.59 t ha^-1, respectively. Hence, the estimated total carbon stock was 152.68±22.95 t ha^-1 and 301.08±27.07 t ha^-1 in degraded and non-degraded community forests, respectively. It was found that the carbon sequestration in the non-degraded community forest was 1.97 times higher than in the degraded community forest. CO₂ equivalent in degraded and non-degraded community forests was 553 t ha^-1 and 1105 t ha^-1, respectively. Statistical analysis showed a significant difference between degraded and non-degraded community forests in terms of its total biomass and carbon sequestration potential (p<0.05). Studies indicate that the community forest has huge potential and can reward economic benefits from carbon trading to benefit from the REDD⁺/CDM mechanism by promoting the sustainable conservation of community forests.