• 전기학회논문지
• /
• 제57권6호
• /
• pp.1058-1062
• /
• 2008

In clinical decision support system(CDSS), unlike rule-based expert method, appropriate data-driven machine learning method can easily provide the information of individual feature(clinical test) for disease classification. However, currently developed methods focus on the improvement of the classification accuracy for diagnosis. With the analysis of feature importance in classification, one may infer the novel clinical test sets which highly differentiate the specific diseases or disease states. In this background, we introduce a novel CDSS that integrate a classifier and feature selection module together. Random forest algorithm is applied for the classifier and the feature importance measure. The system selects the significant clinical tests discriminating the diseases by examining the classification error during backward elimination of the features. The superior performance of random forest algorithm in clinical classification was assessed against artificial neural network and decision tree algorithm by using breast cancer, diabetes and heart disease data in UCI Machine Learning Repository. The test with the same data sets shows that the proposed system can successfully select the significant clinical test set for each disease.

• 한국정보통신학회논문지
• /
• 제10권4호
• /
• pp.687-692
• /
• 2006

기존의 독립 요소 방법에 의한 얼굴인식에서는 주 요소 해석법으로 고유치 크기에 의해 특징을 추출하고 감소된 차원에서 특징 개선을 위한 독립 요소 해석법의 학습을 수행한다. 제거된 특징 공간 내에 필요한 요소가 있는 경우를 고려하지 못한 것이다. 새로운 방법은 독립 요소 해석에 의한 학습을 먼저 시행하고 분리된 데이터를 4차 중심 모멘트에 의한 축적 계수(cumulant)인 커토시스(kurtosis)의 절대값 크기에 의하여 특징을 추출한다. 하지만 독립 요소 방법은 효과적으로 노이즈를 제거하지 못한다. 두 방법의 결합효과는 주 요소 해석법을 노이즈 필터로 사용 할 때 극대화 될 수 있다. 즉 주 요소 해석법을 백색화와 노이즈 필터로 하고 독립 요소 해석법을 특징 추출 방법으로 사용하는 것이다. 실험 결과는 새로운 방법론이 기존의 방법론보다 우수함을 보여준다.

• Journal of Information Processing Systems
• /
• 제14권1호
• /
• pp.176-190
• /
• 2018

This paper presents a novel convolutional neural network based multi-feature fusion learning method for non-rigid 3D model retrieval, which can investigate the useful discriminative information of the heat kernel signature (HKS) descriptor and the wave kernel signature (WKS) descriptor. At first, we compute the 2D shape distributions of the two kinds of descriptors to represent the 3D model and use them as the input to the networks. Then we construct two convolutional neural networks for the HKS distribution and the WKS distribution separately, and use the multi-feature fusion layer to connect them. The fusion layer not only can exploit more discriminative characteristics of the two descriptors, but also can complement the correlated information between the two kinds of descriptors. Furthermore, to further improve the performance of the description ability, the cross-connected layer is built to combine the low-level features with high-level features. Extensive experiments have validated the effectiveness of the designed multi-feature fusion learning method.

• 대한임베디드공학회논문지
• /
• 제16권6호
• /
• pp.323-330
• /
• 2021

Visual object tracking is a challenging area of study in the field of computer vision due to many difficult problems, including a fast variation of target shape, occlusion, and arbitrary ground truth object designation. In this paper, we focus on the reinforced feature of the dynamic search area to get better performance than conventional discriminative model prediction trackers on the condition when the accuracy deteriorates since low feature discrimination. We propose a reinforced input feature method shown like the spotlight effect on the dynamic search area of the target tracking. This method can be used to improve performances for deep learning based discriminative model prediction tracker, also various types of trackers which are used to infer the center of the target based on the visual object tracking. The proposed method shows the improved tracking performance than the baseline trackers, achieving a relative gain of 38% quantitative improvement from 0.433 to 0.601 F-score at the visual object tracking evaluation.

• 한국멀티미디어학회논문지
• /
• 제23권12호
• /
• pp.1540-1551
• /
• 2020

Aiming at the problem that the existing human behavior recognition algorithm cannot fully utilize the multi-level spatio-temporal information of the network, a human behavior recognition algorithm based on a dense three-dimensional residual network is proposed. First, the proposed algorithm uses a dense block of three-dimensional residuals as the basic module of the network. The module extracts the hierarchical features of human behavior through densely connected convolutional layers; Secondly, the local feature aggregation adaptive method is used to learn the local dense features of human behavior; Then, the residual connection module is applied to promote the flow of feature information and reduced the difficulty of training; Finally, the multi-layer local feature extraction of the network is realized by cascading multiple three-dimensional residual dense blocks, and use the global feature aggregation adaptive method to learn the features of all network layers to realize human behavior recognition. A large number of experimental results on benchmark datasets KTH show that the recognition rate (top-l accuracy) of the proposed algorithm reaches 93.52%. Compared with the three-dimensional convolutional neural network (C3D) algorithm, it has improved by 3.93 percentage points. The proposed algorithm framework has good robustness and transfer learning ability, and can effectively handle a variety of video behavior recognition tasks.

• International Journal of Computer Science & Network Security
• /
• 제22권1호
• /
• pp.167-174
• /
• 2022

In metaheuristic algorithms such as Genetic Algorithm (GA), initial population has a significant impact as it affects the time such algorithm takes to obtain an optimal solution to the given problem. In addition, it may influence the quality of the solution obtained. In the machine learning field, feature selection is an important process to attaining a good performance model; Genetic algorithm has been utilized for this purpose by scientists. However, the characteristics of Genetic algorithm, namely random initial population generation from a vector of feature elements, may influence solution and execution time. In this paper, the use of a statistical algorithm has been introduced (Chi2) for feature relevant checks where p-values of conditional independence were considered. Features with low p-values were discarded and subject relevant subset of features to Genetic Algorithm. This is to gain a level of certainty of the fitness of features randomly selected. An ensembled-based learning model for Hepatitis has been developed for Hepatitis C stage classification. 1385 samples were used using Egyptian-dataset obtained from UCI repository. The comparative evaluation confirms decreased in execution time and an increase in model performance accuracy from 56% to 63%.

• 한국문헌정보학회지
• /
• 제53권2호
• /
• pp.179-199
• /
• 2019

본 연구의 목적은 특허 문헌 분류에 가장 적합한 방법론을 발견하기 위하여 다양한 자질 추출 방법과 기계학습 및 딥러닝 모델을 살펴보고 실험을 통해 최적의 성능을 제공하는 방법론을 분석하는데 있다. 자질 추출 방법으로는 전통적인 BoW 방법과 분산표현 방식인 워드 임베딩 벡터를 비교 실험하고, 문헌 집합 구축 방식으로는 형태소 분석과 멀티그램을 이용하는 방식을 비교 검토하였다. 또한 전통적인 기계학습 모델과 딥러닝 모델을 이용하여 분류 성능을 검증하였다. 실험 결과, 분산표현 방법과 형태소 분석을 이용한 자질추출 방법을 기반으로 딥러닝 모델을 적용하였을 경우에 분류 성능이 가장 우수한 것으로 판명되었으며 섹션, 클래스, 서브클래스 분류 실험에서 전통적인 기계학습 방법에 비해 각각 5.71%, 18.84%, 21.53% 우수한 분류 성능을 보여주었다.

• 대한건축학회논문집:구조계
• /
• 제35권11호
• /
• pp.155-162
• /
• 2019

The construction industry is the highest safety accident causing industry as 28.55% portion of all industries' accidents in Korea. In particular, falling is the highest accidents type composed of 60.16% among the construction field accidents. Therefore, we analyzed the factors of major disaster affecting the fall accident and then derived feature importances by considering various variables. We used data collected from Korea Occupational Safety & Health Agency (KOSHA) for learning and predicting in the proposed model. We have an effort to predict the degree of occupational fall accidents by using the machine learning model, i.e., Adaboost, short for Adaptive Boosting. Adaboost is a machine learning meta-algorithm which can be used in conjunction with many other types of learning algorithms to improve performance. Decision trees were combined with AdaBoost in this model to predict and classify the degree of occupational fall accidents. HyOperpt was also used to optimize hyperparameters and to combine k-fold cross validation by hierarchy. We extracted and analyzed feature importances and affecting fall disaster by permutation technique. In this study, we verified the degree of fall accidents with predictive accuracy. The machine learning model was also confirmed to be applicable to the safety accident analysis in construction site. In the future, if the safety accident data is accumulated automatically in the network system using IoT(Internet of things) technology in real time in the construction site, it will be possible to analyze the factors and types of accidents according to the site conditions from the real time data.

• 한국정보통신학회논문지
• /
• 제16권1호
• /
• pp.33-40
• /
• 2012

본 논문은 생물 조기 경보 시스템을 구현하기 위한 방법을 제안한다. 이 시스템은 모니터링 데몬을 이용해 간헐적으로 데이터 사건을 생성하고, 이 데이터 집합으로부터 특징 매개변수들을 추출한다. 특징 매개변수는 6개의 변수(x/y 축 좌표, 거리, 절대 거리, 각도, 프랙털 차원)를 가지고 유도된다. 특히 프랙털 이론을 사용해 제안 알고리즘은 입력된 특징들이 독성 환경에 있는지 아닌지의 유기물 특성을 정의한다. 추출된 특징 데이터를 학습하기 위한 적절한 알고리즘을 위해 기계학습 분야에서 널리 쓰이는 확장된 학습 알고리즘(UChoo)을 사용한다. 그리고 본 알고리즘은 특징 집합들이 모니터링 데몬에 의해 주기적으로 추가된다는 BEWS의 특징을 극복하기 위해 확장된 데이터 표현 방법을 이용하는 학습 방법을 포함한다. 이 알고리즘에서 결정트리 분류기는 확장된 데이터 표현에서 가중치 매개변수를 사용하는 부류 분포 정보를 정의 한다. 실험 결과들은 제안된 BEWS가 환경적인 독성을 탐지하는데 이용 될 수 있음을 보인다.

• 대한원격탐사학회지
• /
• 제29권5호
• /
• pp.509-516
• /
• 2013

초분광 영상에 내재된 비선형 현상을 다루기 위해서는 과거에 주로 사용되었던 선형 피처 추출 방법은 적합하지 않았다. 따라서 최근 Manifold learning이라 불리우는 비선형 피처 추출 방법이 초분광 원격탐사 분야를 비롯 여러 분야에서 관심이 증가되고 있다. Manifold learning 방법 중 널리 이용되는 Isomap은 분류와 분광 혼합 분석 등의 분야에서 좋은 결과를 보여주지만, 지나치게 복잡하고 높은 계산량은, 특히 원격탐사 자료와 같이 자료의 크기가 큰 경우 문제가 된다. 따라서 자료의 일부분을 이용하는 랜드마크 기법이 해결책으로 제안 되었다. 본 연구에서는 좀 더 통제가 가능한 랜드마크 추출을 위해 자료를 구성하는 최대 부피를 지닌 Simplex를 이용하여 랜드마크를 선택하는 방법을 제안한다. 초분광 영상을 이용하여 랜드마크의 개수, 선택 방법에 따른 분류 정확도와 편차, 그리고 처리 시간을 비교하였고, 그 결과 제안된 랜드마크 선택 기법은 분류 정확도, 처리시간 모두에서 효율적인 결과를 보여주었다.