• Journal of the Korean Wood Science and Technology
• /
• 제47권3호
• /
• pp.265-276
• /
• 2019

In our previous study, the LeNet3 model successfully classified images from the transverse surfaces of five Korean softwood species (cedar, cypress, Korean pine, Korean red pine, and larch). However, a practical limitation exists in our system stemming from the nature of the training images obtained from the transverse plane of the wood species. In real-world applications, it is necessary to utilize images from the longitudinal surfaces of lumber. Thus, we improved our model by training it with images from the longitudinal and transverse surfaces of lumber. Because the longitudinal surface has complex but less distinguishable features than the transverse surface, the classification performance of the LeNet3 model decreases when we include images from the longitudinal surfaces of the five Korean softwood species. To remedy this situation, we adopt ensemble methods that can enhance the classification performance. Herein, we investigated the use of ensemble models from the LeNet and MiniVGGNet models to automatically classify the transverse and longitudinal surfaces of the five Korean softwoods. Experimentally, the best classification performance was achieved via an ensemble model comprising the LeNet2, LeNet3, and MiniVGGNet4 models trained using input images of $128{\times}128{\times}3pixels$ via the averaging method. The ensemble model showed an F1 score greater than 0.98. The classification performance for the longitudinal surfaces of Korean pine and Korean red pine was significantly improved by the ensemble model compared to individual convolutional neural network models such as LeNet3.

• 정보보호학회논문지
• /
• 제32권2호
• /
• pp.267-277
• /
• 2022

최근 클라우드 및 원격 근무 환경의 비중이 증가함에 따라 다양한 정보보안 사고들이 발생하고 있다. 조직의 내부자가 원격 접속으로 기밀 자료에 접근하여 유출을 시도하는 사례가 발생하는 등 내부자 위협이 주요 이슈로 떠오르게 되었다. 이에 따라 내부자 위협을 탐지하기 위해 기계학습 기반의 방법들이 제안되고 있다. 하지만, 기존의 내부자 위협을 탐지하는 기계학습 기반의 방법들은 편향 및 분산 문제와 같이 예측 정확도와 관련된 중요한 요소를 고려하지 않았으며 이에 따라 제한된 성능을 보인다는 한계가 있다. 본 논문에서는 편향 및 분산을 고려하는 부스팅 유형의 앙상블 학습 알고리즘들을 사용하여 악의적인 내부자 탐지 성능을 확인하고 이에 대한 면밀한 분석을 수행하며, 데이터셋의 불균형까지도 고려하여 최종 결과를 판단한다. 앙상블 학습을 이용한 실험을 통해 기존의 단일 학습 모델에 기반한 방법에서 나아가, 편향-분산 트레이드오프를 함께 고려하며 유사하거나 보다 높은 정확도를 달성함을 보인다. 실험 결과에 따르면 배깅과 부스팅 방법을 사용한 앙상블 학습은 98% 이상의 정확도를 보였고, 이는 사용된 단일 학습 모델의 평균 정확도와 비교하면 악의적인 내부자 탐지 성능을 5.62% 향상시킨다.

• Computers and Concrete
• /
• 제32권3호
• /
• pp.313-326
• /
• 2023

The seismic safety of the shear wall structure can be assessed through seismic fragility analysis, which requires high computational costs in estimating seismic demands. Accordingly, machine learning methods have been applied to such fragility analyses in recent years to reduce the numerical analysis cost, but it still remains a challenging task. Therefore, this study uses the ensemble machine learning method to present an improved framework for developing a more accurate seismic demand model than the existing ones. To this end, a rank-based selection method that enables determining an excellent model among several single machine learning models is presented. In addition, an index that can evaluate the degree of overfitting/underfitting of each model for the selection of an excellent single model is suggested. Furthermore, based on the selected single machine learning model, we propose a method to derive a more accurate ensemble model based on the bagging method. As a result, the seismic demand model for which the proposed framework is applied shows about 3-17% better prediction performance than the existing single machine learning models. Finally, the seismic fragility obtained from the proposed framework shows better accuracy than the existing fragility methods.

• 한국소음진동공학회논문집
• /
• 제26권6_spc호
• /
• pp.736-743
• /
• 2016

This study describes optimization methods for adjustment of helicopter main rotor tracking and balancing (RTB). RTB is a essential process for helicopter operation and maintenance. Linear and non-linear models were developed with past RTB test results for estimation of RTB adjustment. Then global and sequential optimization methods were applied to the each of models. Utilization of the individual optimization method with each model is hard to fulfill the RTB requirements because of different characteristics of each blade. Therefore an ensemble model was used to integrate every estimated adjustment result, and an adaptive method was also applied to adjustment values of the linear model to update for next estimations. The goal of this developed RTB adjustment optimization program is to achieve the requirements within 2 run. Additional tests for comparison of weight factor of the ensemble model are however necessary.

• Communications for Statistical Applications and Methods
• /
• 제29권6호
• /
• pp.629-640
• /
• 2022

Variable selection is one of the most crucial tasks in supervised learning, such as regression and classification. The best subset selection is straightforward and optimal but not practically applicable unless the number of predictors is small. In this article, we propose directly solving the best subset selection via the genetic algorithm (GA), a popular stochastic optimization algorithm based on the principle of Darwinian evolution. To further improve the variable selection performance, we propose to run multiple GA to solve the best subset selection and then synthesize the results, which we call ensemble GA (EGA). The EGA significantly improves variable selection performance. In addition, the proposed method is essentially the best subset selection and hence applicable to a variety of models with different selection criteria. We compare the proposed EGA to existing variable selection methods under various models, including linear regression, Poisson regression, and Cox regression for survival data. Both simulation and real data analysis demonstrate the promising performance of the proposed method.

• 예술인문사회 융합 멀티미디어 논문지
• /
• 제8권3호
• /
• pp.57-67
• /
• 2018

앙상블 기법은 기계학습에서 다수의 알고리즘을 사용하여 더 좋은 성능을 내기 위해 사용하는 방법이다. 본 논문에서는 앙상블 기법에서 많이 사용되는 부스팅과 배깅에 대해 소개를 하고, 서포트벡터 회귀, 방사기저함수 네트워크, 가우시안 프로세스, 다층 퍼셉트론을 이용하여 설계한다. 추가적으로 순환신경망과 MOHID 수치모델을 추가하여 실험을 진행한다. 실험적 검증를 위해 사용하는 뜰개 데이터는 7 개의 지역에서 관측된 683 개의 관측 자료다. 뜰개 관측 자료를 이용하여 6 개의 알고리즘과의 비교를 통해 앙상블 기법의 성능을 검증한다. 검증 방법으로는 평균절대오차를 사용한다. 실험 방법은 배깅, 부스팅, 기계학습을 이용한 앙상블 모델을 이용하여 진행한다. 각 앙상블 모델마다 동일한 가중치를 부여한 방법, 차등한 가중치를 부여한 방법을 이용하여 오류율을 계산한다. 가장 좋은 오류율을 나타낸 방법은 기계학습을 이용한 앙상블 모델로서 6 개의 기계학습의 평균에 비해 61.7%가 개선된 결과를 보였다.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2007년도 추계학술대회논문집
• /
• pp.975-978
• /
• 2007

This paper investigates stage acoustics in concert halls for ensemble performance in terms of stage support parameters. A concert hall with large stage area was selected. Objective measurements were carried out according to Gade's methods for investigating stage support (ST1). The sound field on the stage was evaluated using stage and audience acoustical parameters. According to the positions at stage, different acoustical characteristics were found. Higher ST1 and lower $IACC_{E3}$ were measured near side and rear stage walls. In case of conductor/soloist vs. orchestra player, objective measurement of acoustical parameters shows that early sound fields depend on the source-receiver distance but late sound fields depend on the location of side and rear walls. In addition, in-situ evaluation of the trio instrumentalists was carried out. Results show that ensemble is favored by higher early reflection of sounds and sound diffusion affecting ‘blending’ of their performance.

• International Journal of Internet, Broadcasting and Communication
• /
• 제10권4호
• /
• pp.50-64
• /
• 2018

This paper describes a visual object detection approach utilizing ensemble based machine learning. Object detection methods employing 1D features have the benefit of fast calculation speed. However, for real image with complex background, detection accuracy and performance are degraded. In this paper, we propose an ensemble learning algorithm that combines a 1D feature classifier and 2D DNF (Disjunctive Normal Form) classifier to improve the object detection performance in a single input image. Also, to improve the computing efficiency and accuracy, we propose a feature selecting method to reduce the computing time and ensemble algorithm by combining the 1D features and 2D DNF features. In the verification experiments, we selected the Haar-like feature as the 1D image descriptor, and demonstrated the performance of the algorithm on a few datasets such as face and vehicle.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권11호
• /
• pp.5592-5609
• /
• 2017

JPEG steganography detection is an active research topic in the field of information hiding due to the wide use of JPEG image in social network, image-sharing websites, and Internet communication, etc. In this paper, a new steganalysis method for content-adaptive JPEG steganography is proposed by integrating the evolutionary feature selection and classifier ensemble selection. First, the whole framework of the proposed steganalysis method is presented and then the characteristic of the proposed method is analyzed. Second, the feature selection method based on genetic algorithm is given and the implement process is described in detail. Third, the method of classifier ensemble selection is proposed based on Pareto evolutionary optimization. The experimental results indicate the proposed steganalysis method can achieve a competitive detection performance by compared with the state-of-the-art steganalysis methods when used for the detection of the latest content-adaptive JPEG steganography algorithms.

• ETRI Journal
• /
• 제41권6호
• /
• pp.750-759
• /
• 2019

Presently, the extraction of hand-crafted features is still the dominant method in radar emitter recognition. To solve the complicated problems of selection and updation of empirical features, we present a novel automatic feature extraction structure based on deep learning. In particular, a convolutional neural network (CNN) is adopted to extract high-level abstract representations from the time-frequency images of emitter signals. Thus, the redundant process of designing discriminative features can be avoided. Furthermore, to address the performance degradation of a single platform, we propose the construction of an ensemble learning-based architecture for multi-platform fusion recognition. Experimental results indicate that the proposed algorithms are feasible and effective, and they outperform other typical feature extraction and fusion recognition methods in terms of accuracy. Moreover, the proposed structure could be extended to other prevalent ensemble learning alternatives.