• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.21 no.11
• /
• pp.1020-1028
• /
• 2011

As rotating machines play an important role in industrial applications such as aeronautical, naval and automotive industries, many researchers have developed various condition monitoring system and fault diagnosis system by applying artificial neural network. Since using obtained signals without preprocessing as inputs of neural network can decrease performance of fault classification, it is very important to extract significant features of captured signals and to apply suitable features into diagnosis system according to the kinds of obtained signals. Therefore, this paper proposes a neural-network-based fault diagnosis system using AR coefficients as feature vectors by LPC(linear predictive coding) and EIV(errors-in variables) analysis. We extracted feature vectors from sound, vibration and current faulty signals and evaluated the suitability of feature vectors depending on the classification results and training error rates by changing AR order and adding noise. From experimental results, we conclude that classification results using feature vectors by EIV analysis indicate more than 90 % stably for less than 10 orders and noise effect comparing to LPC.

• Journal of the Korean Society for information Management
• /
• v.18 no.2
• /
• pp.203-230
• /
• 2001

The purpose of this study is to develop a document clustering model for automatic classification of knowledge. Two test collections of newspaper article texts and journal article abstracts are built for the clustering experiment. Various feature reduction criteria as well as term weighting methods are applied to the term sets of the test collections, and cosine and Jaccard coefficients are used as similarity measures. The performances of complete linkage and K-means clustering algorithms are compared using different feature selection methods and various term weights. It was found that complete linkage clustering outperforms K-means algorithm and feature reduction up to almost 10% of the total feature sets does not lower the performance of document clustering to any significant extent.

• Journal of Biomedical Engineering Research
• /
• v.18 no.4
• /
• pp.357-364
• /
• 1997

The early detection of breast cancer is clearly a key ingredient for reducing breast cancer mortality. Microcalcification is the only visible feature of the DCIS's(ductal carcinoma in situ) which consist 15 ~ 20% of screening-detected breast cancer. Therefore, the analysis of the shapes and distributions of microcalcifications is very significant for the early detection. The automatic detection procedures have b(:on the concern of digital image processing for many years. We proposed here one efficient method which is essentially statistical pattern classification accelerated by one representative feature, correlation coefficient. We compared the results by this additional feature with results by a simple gray level thresholding. The average detection rate was increased from 48% by gray level feature only to 83% by the proposed method The performances were evaluated with TP rates and FP counts, and also with Bayes errors.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.20 no.1
• /
• pp.33-39
• /
• 2017

In this paper, an effective feature which is capable of classifying targets among the detections obtained from 2D range-bearing maps generated in active sonar environments is proposed. Most conventional approaches for target classification with the 2D maps have considered magnitude of peak and statistical features of the area surrounding the peak. To improve the classification performance, HOG(Histogram of Gradient) feature, which is popular for their robustness in the image textures analysis is applied. In order to classify the target signal, SVM(Support Vector Machine) method with reduced HOG feature by the PCA(Principal Component Analysis) algorithm is incorporated. The various simulations are conducted with the real clutter signal data and the synthesized target signal data. According to the simulated results, the proposed method considering HOG feature is claimed to be effective when classifying the active sonar target compared to the conventional methods.

• Journal of Korea Society of Industrial Information Systems
• /
• v.18 no.4
• /
• pp.25-35
• /
• 2013

This paper focuses on building a generalized mood classification model with many mood classes instead of a personalized one with few mood classes. Two methods are adopted to improve the performance of mood classification. The one of them is feature reduction based on standard deviation of feature values, which is designed to solve the problem of lowered performance when all 391 features provided by MIR toolbox used to extract features of music. The experiments show that the feature reduction methods suggested in this paper have better performance than that of the conventional dimension reduction methods, R-Square and PCA. As performance improvement by feature reduction only is subject to limit, modular neural network is used as another method to improve the performance. The experiments show that the method also improves performance effectively.

• Korean Journal of Remote Sensing
• /
• v.37 no.4
• /
• pp.719-731
• /
• 2021

This study proposes a two-stage hybrid classification model for crop classification using multi-temporal remote sensing images; the model combines feature embedding by using an autoencoder (AE) with a convolutional neural network (CNN) classifier to fully utilize features including informative temporal and spatial signatures. Long short-term memory (LSTM)-based AE (LAE) is fine-tuned using class label information to extract latent features that contain less noise and useful temporal signatures. The CNN classifier is then applied to effectively account for the spatial characteristics of the extracted latent features. A crop classification experiment with multi-temporal unmanned aerial vehicle images is conducted to illustrate the potential application of the proposed hybrid model. The classification performance of the proposed model is compared with various combinations of conventional deep learning models (CNN, LSTM, and convolutional LSTM) and different inputs (original multi-temporal images and features from stacked AE). From the crop classification experiment, the best classification accuracy was achieved by the proposed model that utilized the latent features by fine-tuned LAE as input for the CNN classifier. The latent features that contain useful temporal signatures and are less noisy could increase the class separability between crops with similar spectral signatures, thereby leading to superior classification accuracy. The experimental results demonstrate the importance of effective feature extraction and the potential of the proposed classification model for crop classification using multi-temporal remote sensing images.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.6
• /
• pp.2806-2825
• /
• 2018

Unsupervised learning has shown good performance on image, video and audio classification tasks, and much progress has been made so far. It studies how systems can learn to represent particular input patterns in a way that reflects the statistical structure of the overall collection of input patterns. Many promising deep learning systems are commonly trained by the greedy layerwise unsupervised learning manner. The performance of these deep learning architectures benefits from the unsupervised learning ability to disentangling the abstractions and picking out the useful features. However, the existing unsupervised learning algorithms are often difficult to train partly because of the requirement of extensive hyperparameters. The tuning of these hyperparameters is a laborious task that requires expert knowledge, rules of thumb or extensive search. In this paper, we propose a simple and effective unsupervised feature learning algorithm for image classification, which exploits an explicit optimizing way for population and lifetime sparsity. Firstly, a sparse target matrix is built by the competitive rules. Then, the sparse features are optimized by means of minimizing the Euclidean norm ($L_2$) error between the sparse target and the competitive layer outputs. Finally, a classifier is trained using the obtained sparse features. Experimental results show that the proposed method achieves good performance for image classification, and provides discriminative features that generalize well.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.13 no.4
• /
• pp.91-98
• /
• 2017

Previous works for detecting arrhythmia have mostly used nonlinear method such as artificial neural network, fuzzy theory, support vector machine to increase classification accuracy. Most methods require higher computational cost and larger processing time. Therefore it is necessary to design efficient algorithm that classifies PVC(premature ventricular contraction) and decreases computational cost by accurately detecting minimal feature point based on only R peak through optimal R wave. We propose an optimal R wave detection and PVC classification method through extracting minimal feature point in IoT environment. For this purpose, we detected R wave through optimal threshold value and extracted RR interval and R peak pattern from noise-free ECG signal through the preprocessing method. Also, we classified PVC in realtime through RR interval and R peak pattern. The performance of R wave detection and PVC classification is evaluated by using record of MIT-BIH arrhythmia database. The achieved scores indicate the average of 99.758% in R wave detection and the rate of 93.94% in PVC classification.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.10 no.5
• /
• pp.133-138
• /
• 2010

In this paper, we make up a feature image including spatial properties and statistical properties on image, and format covariance matrices using region variance magnitudes. By using it to texture classification, this paper puts a proposal for tough texture classification way to illumination, noise and rotation. Also we offer a way to minimalize performance time of texture classification using integral image expressing middle image for fast calculation of region sum. To estimate performance evaluation of proposed way, this paper use a Brodatz texture image, and so conduct a noise addition and histogram specification and create rotation image. And then we conduct an experiment and get better performance over 96%.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.6
• /
• pp.1636-1645
• /
• 1996

In this paper we proposed a new scheme for conditional two dimensional (2-D)moment-based classification of patterns on the basis of Gibbs random fields which are will suited for representing spatial continuity that is the characteristic of the most images. This implementation contains two parts: feature extraction and pattern classification. First of all, we extract feature vector which consists of conditional 2-D moments on the basis of estimated Gibbs parameter. Note that the extracted feature vectors are invariant under translation, rotation, size of patterns the corresponding template pattern. In order to evaluate the performance of the proposed scheme, classification experiments with training document sets of characters have been carried out on 486 66Mhz PC. Experiments reveal that the proposed scheme has high classification rate over 94%.