• 한국방사선학회논문지
• /
• 제15권5호
• /
• pp.613-620
• /
• 2021

본 논문에서는 흉부 X선 영상에서 정상 심장과 비정상 심장(심장비대)을 분류할 수 있는 합성곱 신경망 모델을 제안하고자 한다. 학습 및 테스트 데이터로는 경북대학교병원에 내원하여 정상과 심장비대를 진단받은 환자들의 흉부 X-선 이미지를 획득하여 사용하였다. 제안된 합성곱 신경망 모델을 이용하였을 때의 정상 심장 및 비정상 심장(심장비대) 분류 정확도는 99.88%였다. 정상 심장 영상을 테스트 데이터로 사용하였을 때의 정확도, 정밀도, 재현율 및 F1 Score는 95%, 100%, 90%, 96%였다. 비정상 심장(심장비대) 영상을 테스트 데이터로 사용하였을 때의 정확도, 정밀도, 재현율 및 F1 Score는 95%, 92%, 100% 및 96%였다. 이러한 학습 및 테스트 분류 결과로 제안된 합성곱 신경망 모델은 흉부 X-선 영상의 특징 추출 및 분류에서 매우 우수한 성능을 보여주고 있다고 판단된다. 본 논문에서 제안하는 합성곱 신경망 모델은 흉부 X-선 영상의 질환 분류에 있어 유용한 결과를 보여줄 것으로 판단되며, 다른 의료 영상에서도 동일한 결과를 나타내는지 알아보기 위하여 추가적인 연구가 이루어져야 할 것이다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제11권3호
• /
• pp.204-210
• /
• 2011

This paper proposes a fuzzy support vector machine ($FSVM_n$) pattern classifier to classify the arrhythmia patterns of an electrocardiograph (ECG). The $FSVM_n$ is a pattern classifier which combines n-dimensional fuzzy membership functions with a slack variable of SVM. To evaluate the performance of the proposed classifier, the MIT/BIH ECG database, which is a standard database for evaluating arrhythmia detection, was used. The pattern classification experiment showed that, when classifying ECG into four patterns - NSR, VT, VF, and NSR, VT, and VF classification rate resulted in 99.42%, 99.00%, and 99.79%, respectively. As a result, the $FSVM_n$ shows better pattern classification performance than the existing SVM and FSVM algorithms.

• Communications for Statistical Applications and Methods
• /
• 제26권6호
• /
• pp.591-610
• /
• 2019

Deep Learning is one of the machine learning methods to find features from a huge data using non-linear transformation. It is now commonly used for supervised learning in many fields. In particular, Convolutional Neural Network (CNN) is the best technique for the image classification since 2012. For users who consider deep learning models for real-world applications, Keras is a popular API for neural networks written in Python and also can be used in R. We try examine the parameter estimation procedures of Deep Neural Network and structures of CNN models from basics to advanced techniques. We also try to figure out some crucial steps in CNN that can improve image classification performance in the CIFAR10 dataset using Keras. We found that several stacks of convolutional layers and batch normalization could improve prediction performance. We also compared image classification performances with other machine learning methods, including K-Nearest Neighbors (K-NN), Random Forest, and XGBoost, in both MNIST and CIFAR10 dataset.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2003년도 봄 학술발표논문집 Vol.30 No.1 (B)
• /
• pp.178-180
• /
• 2003

EEG-based brain computer interface(BCI) provides a new communication channel between human brain and computer. EEG data is a multivariate time series so that hidden Markov model (HMM) might be a good choice for classification. However EEG is very noisy data and contains artifacts, so useful features mr expected to improve the performance of HMM. In this paper we addresses the usefulness of principal component features with Hidden Markov model (HHM). We show that some selected principal component features can suppress small noises and artifacts, hence improves classification performance. Experimental study for the classification of EEG data during imagination of a left, right up or down hand movement confirms the validity of our proposed method.

• Journal of information and communication convergence engineering
• /
• 제15권3호
• /
• pp.182-186
• /
• 2017

Maintaining a proper lookout activity routine is integral to preventing ship collision accidents caused by human errors. Various subjective measures such as interviewing, self-report diaries, and questionnaires have been widely used to monitor the lookout activity patterns of navigators. An objective measurement of a lookout activity pattern classification system is required to improve lookout performance evaluation in a real navigation setting. The purpose of this study was to develop an objective navigator lookout activity classification system using wearable accelerometers. In the training session, 90.4% accuracy was achieved in classifying five fundamental lookout activities. The developed model was then applied to predict real-lookout activity in the second session during an actual ship voyage. 86.9% agreement was attained between the directly observed activity and predicted activity. Based on these promising results, the proposed unobstructed wearable system is expected to objectively evaluate navigator lookout patterns to provide a better understanding of lookout performance.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회 1996년도 추계학술대회발표논문집; 고려대학교, 서울; 26 Oct. 1996
• /
• pp.31-34
• /
• 1996

The extensive researches have compared the performance of neural networks(NN) with those of various statistical techniques for the classification problem. The empirical results of these comparative studies have indicated that the neural networks often outperform the traditional statistical techniques. Moreover, there are some efforts that try to combine various classification methods, especially multivariate discriminant analysis with neural networks. While these efforts improve the performance, there exists a problem violating robust assumptions of multivariate discriminant analysis that are multivariate normality of the independent variables and equality of variance-covariance matrices in each of the groups. On the contrary, cluster analysis alleviates this assumption like neural networks. We propose a new approach to classification problems by combining the cluster analysis with neural networks. The resulting predictions of the composite model are more accurate than each individual technique.

• 전기학회논문지
• /
• 제56권7호
• /
• pp.1224-1230
• /
• 2007

In this paper, we study the classification of defects at steam generator tube in nuclear power plant using eddy current testing (ECT). We consider 4 defect patterns of SG tube: I-In type, I-Out type, V-In type, and V-Out type. Through numerical analysis program based on finite element modeling, 400 ECT signals are generated by varying width and depth of each defect type. In order to improve the classification performance, we propose new feature extraction technique. After extracting new features from the generated ECT signals, multi-layer perceptron is used to classify the defect patterns. Through the computer simulation study, it is shown that the proposed method achieves 100% classification success rate while the previous method yields 91% success rate.

• 인지과학
• /
• 제10권1호
• /
• pp.13-24
• /
• 1999

논문에서는 모듈화 신경 을 이용한 비지도 학습방법의 분류기를 제안한다. 각 모듈은 데이터의 통계학적인 분석의 결과로 설계되어져서, 데이터의 독립적인 군집들을 나타내게 된다. 이런 신경의 독립적인 분류 결과와 근접거리 척도를 이용한 유사도 측정을 통해 더욱 정확한 분류를 가능케 하며, 오 분류를 하는 모듈을 삭제함으로써 계산 을 줄인다. 이런 과정을 통해 신경 에 사용되는 각종 변수에 대한 별다른 조사 과정 없이 최상의 성능을 발휘하는 신경 에 준 는 성능을 가진 신경 망을 구축했다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.171-171
• /
• 2000

We have developed m innovative surface inspection system for automated quality control for steel products in POSCO. We had ever installed the various kinds of surface inspection systems, such as a linear CCD and a laser typed surface inspection systems at cold rolled strips production lines. But, these systems cannot fulfill the sufficient detection and classification rate, and real time processing performance. In order to increase detection and classification rate, we have used the Dark, Bright and Transition Field illumination and area type CCD camera, and fur the real time image processing, parallel computing has been used. In this paper, we introduced the automatic surface inspection system and real time image processing technique using the Object Detection, Defect Detection, Classification algorithms and its performance obtained at the production line.

• 대한의용생체공학회:의공학회지
• /
• 제21권2호
• /
• pp.137-144
• /
• 2000

Clustered microcalcifications in X-ray mammograms are an important sign for the diagnosis of breast cancer. A shape-based method, which is based on the morphological features of clustered microcalcifications, is proposed for classifying clustered microcalcifications into benign or malignant categories. To verify the effectiveness of the proposed shape features, clinical mammograms were used to compare the classification performance of the proposed shape features with those of conventional textural features, such as the spatial gray-leve dependence method and the wavelet-based method. Image features extracted from these methods were used as inputs to a three-layer backpropagation neural network classifier. The classification performance of features extracted by each method was studied by using receiver operating-characteristics analysis. The proposed shape features were shown to be superior to the conventional textural features with respect to classification accuracy.