• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.9
• /
• pp.2927-2941
• /
• 2022

The primary task of machine fault diagnosis is to judge whether the current state is normal or damaged, so it is a typical binary classification problem with mutual exclusion. Mutually exclusive events and out-of-domain detection have one thing in common: there are two types of data and no intersection. We proposed a fusion model method to improve the accuracy of machine fault diagnosis, which is based on the mutual exclusivity of events and the commonality of out-of-distribution detection, and finally generalized to all binary classification problems. It is reported that the performance of a convolutional neural network (CNN) will decrease as the recognition type increases, so the variational auto-encoder (VAE) is used as the primary model. Two VAE models are used to train the machine's normal and fault sound data. Two reconstruction probabilities will be obtained during the test. The smaller value is transformed into a correction value of another value according to the mutually exclusive characteristics. Finally, the classification result is obtained according to the fusion algorithm. Filtering normal data features from fault data features is proposed, which shields the interference and makes the fault features more prominent. We confirm that good performance improvements have been achieved in the machine fault detection data set, and the results are better than most mainstream models.

• Asian Pacific Journal of Cancer Prevention
• /
• v.17 no.2
• /
• pp.835-838
• /
• 2016

Background: Lung cancer remains one of the most common cancers in the world, both in terms of new cases (about 13% of total per year) and deaths (nearly one cancer death in five), because of the high case fatality. Errors in lung cancer type or malignant growth determination lead to degraded treatment efficacy, because anticancer strategy depends on tumor morphology. Materials and Methods: We have made an attempt to evaluate effectiveness of machine learning algorithms in the task of lung cancer classification based on gene expression levels. We processed four publicly available data sets. The Dana-Farber Cancer Institute data set contains 203 samples and the task was to classify four cancer types and sound tissue samples. With the University of Michigan data set of 96 samples, the task was to execute a binary classification of adenocarcinoma and non-neoplastic tissues. The University of Toronto data set contains 39 samples and the task was to detect recurrence, while with the Brigham and Women's Hospital data set of 181 samples it was to make a binary classification of malignant pleural mesothelioma and adenocarcinoma. We used the k-nearest neighbor algorithm (k=1, k=5, k=10), naive Bayes classifier with assumption of both a normal distribution of attributes and a distribution through histograms, support vector machine and C4.5 decision tree. Effectiveness of machine learning algorithms was evaluated with the Matthews correlation coefficient. Results: The support vector machine method showed best results among data sets from the Dana-Farber Cancer Institute and Brigham and Women's Hospital. All algorithms with the exception of the C4.5 decision tree showed maximum potential effectiveness in the University of Michigan data set. However, the C4.5 decision tree showed best results for the University of Toronto data set. Conclusions: Machine learning algorithms can be used for lung cancer morphology classification and similar tasks based on gene expression level evaluation.

• Journal of Korean Clinical Nursing Research
• /
• v.15 no.1
• /
• pp.5-17
• /
• 2009

Purpose: This study was to develop a factor-type patient classification system for general nursing unit based on nursing needs (KPCS; Korean patient classification system for nurses). Method: We reviewed workload management system for nurses(WMSN) of Walter Reed Medical Center, Korean patient classification system for ICU, and nursing activities in nursing records and developed the first version of KPCS. The final version KPCS was evaluated via validity and reliability verifications based on panel discussions and data from 800 patient classifications. Content validity was performed by Delphi method and concurrent validity was verified by the correlation of two tools (r=.71). Construct validity was also tested by medical department (p<.001), patient type (p<.001), and nurse intuition (p<.001). These verifications were performed from April to October, 2008. Results: The KPCS has 75 items in classifying 50 nursing activities, and categorized into 12 different nursing area (measuring vital sign, monitoring, respiratory treatment, hygiene, diet, excretion, movement, examination, medication, treatment, special treatment, and education/emotional support). Conclusion: The findings of the study showed sound reliability and validity of KPCS based on nursing needs. Further study is mandated to refine the system and to develop index score to estimate the necessary number of nurses for adequate care.

• Journal of the Korean Wood Science and Technology
• /
• v.23 no.2
• /
• pp.47-54
• /
• 1995

This study was to measure the properties of logs and classify them by non-destructive methods. The purpose of this experiment was focused at mensurability of logs by non-destructive methods. The non-destructive instrument, Stress-Wave Timer 239A which was made by Metriguard in U.S.A., was used. The stress wave velocities of log's cross direction were measured and compared with three different methods; 1. with hammer, 2. with hammer and D.B.H. meter, 3. with manufactured instrument. Number of used logs were seven logs, which were classified by naked eye into six groups; very severe rot, severe rot, mild rot & knot, mild rot & check, mild rot, sound log, and in diameter were into three groups; large(57.4cm), medium(36~41.2cm), small(28.9cm) log. The results, which were classified by mensurability with non-destructive methods, were followed; 1. The stress wave velocities were very different between rot and sound log. So it meant the possibility of mensurability of logs by non-destructive method even if high standard error. 2. The stress wave velocities decreased with checks more than with rots, which meant the checks affected speeds more. 3. The stress wave velocities increased with knot. 4. The velocities with manufactured instrument showed lower standard error, so more accurate results than other methods. Especially the required labour decreased from 3~4 to 2 persons. 5. Finally, the mensurability showed more accurate results and made the classification of logs scientific.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.3
• /
• pp.312-318
• /
• 2022

This paper proposes a model which classifies the type of guns and information about sound source location using deep neural network. The proposed classification model is composed of convolutional neural networks (CNN) and long short-term memory (LSTM). For training and test the model, we use the Gunshot Audio Forensic Dataset generated by the project supported by the National Institute of Justice (NIJ). The acoustic signals are transformed to Mel-Spectrogram and they are provided as learning and test data for the proposed model. The model is compared with the control model consisting of convolutional neural networks only. The proposed model shows high accuracy more than 90 %.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.21 no.6
• /
• pp.183-190
• /
• 2011

In this paper, we propose a prototype system for abnormal sound detection and identification which detects and recognizes the abnormal situations by means of analyzing audio information coming in real time from CCTV cameras under surveillance environment. The proposed system is composed of two layers: The first layer is an one-class support vector machine, i.e., support vector data description (SVDD) that performs rapid detection of abnormal situations and alerts to the manager. The second layer classifies the detected abnormal sound into predefined class such as 'gun', 'scream', 'siren', 'crash', 'bomb' via a sparse representation classifier (SRC) to cope with emergency situations. The proposed system is designed in a hierarchical manner via a mixture of SVDD and SRC, which has desired characteristics as follows: 1) By fast detecting abnormal sound using SVDD trained with only normal sound, it does not perform the unnecessary classification for normal sound. 2) It ensures a reliable system performance via a SRC that has been successfully applied in the field of face recognition. 3) With the intrinsic incremental learning capability of SRC, it can actively adapt itself to the change of a sound database. The experimental results with the qualitative analysis illustrate the efficiency of the proposed method.

• Proceedings of the Korean Society for Emotion and Sensibility Conference
• /
• 2002.05a
• /
• pp.24-29
• /
• 2002

The purpose of this study was to classify apparel fabrics according to rustling sounds and to analyze their transformed colors and mechanical properties. The rustling sounds of apparel fabrics were recorded and then transformed into colors using Mori's color-transforming program. The specimens were clustered into five groups according to sound properties, and each group was named as 'Silky', 'Crispy', 'Paper-like', 'Worsted', and 'Flaxy', respectively. The Silky consisted of smooth and soft silk fabrics had the lowest value of LPT, $\Delta$f ARC, loudness(z) and sharpness(z). Their transformed colors showed lots of red portion and color counts. The Crispy with crepe fabrics showed relatively low loudness(z) and sharpness(z), but diverse colors and color counts were appeared. The Paper-like showed the highest value of LPT, $\Delta$f and loudness(z). The Worsted composed of wool and wool-like fabrics showed high values of LPT, $\Delta$f loudness(z) and sharpness(z). The transformed colors of the Paper-like and Worsted showed the blue mostly but color counts were less than the others. The Flaxy with rugged flax fabric had the highest fluctuation strength, and their transformed colors showed diversity.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.07a
• /
• pp.18-21
• /
• 2020

심층 신경망은 영상 분류, 음성 인식, 그리고 문자 번역 등 다양한 분야에서 효과적인 성능을 보여주고 있다. 신경망의 구조 변화, 신경망 간의 정보 전달, 그리고 학습에 사용되는 데이터 증대 등의 확장된 연구를 통해 성능은 더욱 발전하고 있다. 그 중에서도 데이터 증대는 기존에 수집한 데이터의 변형을 통해 심층 신경망에 더 다양한 데이터를 제공함으로써 더욱 일반화된 신경망을 학습시기키는 것을 목표로 한다. 하지만 기존의 음향 관련 신경망 연구에서는 모델의 학습에 사용되는 데이터 증대 방법의 연구가 영상 처리 분야만큼 다양하게 이루어지지 않았다. 최근 영상 처리 분야의 데이터 증대 연구는 학습에 사용되는 데이터와 모델에 따라 최적의 데이터 증대 방법이 다르다는 것을 실험적으로 보여주었다. 이에 영감을 받아 본 논문은 자연에서 발생하는 음향을 분류하는데 있어서 최적의 데이터 증대 방법을 실험적으로 찾으며, 그 과정을 소개한다. 음향에 잡음 추가, 피치 변경 혹은 스펙트로그램의 일부 제한 등의 데이터 증대 방법을 다양하게 조합하는 실험을 통해 경험적으로 어떤 증대 방법이 효과적인지 탐색했다. 결과적으로 ESC-50 자연 음향 데이터 셋에 최적화된 데이터 증대 방법을 적용함으로써 분류 정확도를 89%로 향상시킬 수 있었다.

• The Journal of the Acoustical Society of Korea
• /
• v.23 no.5
• /
• pp.400-409
• /
• 2004

In this paper. a voice personality transformation method is proposed. which makes one person's voice sound like another person's voice. To transform the voice personality. vocal tract transfer function is used as a transformation parameter. Comparing with previous methods. the proposed method makes transformed speech closer to target speaker's voice in both subjective and objective points of view. Conversion between vocal tract transfer functions is implemented by classification of entire vector space followed by linear transformation for each cluster. LPC cepstrum is used as a feature parameter. A joint classification and transformation method is proposed, where optimum clusters and transformation matrices are simultaneously estimated in the sense of a minimum mean square error criterion. To evaluate the performance of the proposed method. transformation rules are generated from 150 sentences uttered by three male and on female speakers. These rules are then applied to another 150 sentences uttered by the same speakers. and objective evaluation and subjective listening tests are performed.

• Journal of the Institute of Convergence Signal Processing
• /
• v.18 no.2
• /
• pp.43-48
• /
• 2017

The underwater transients signals contain the characteristics of complexity, time varying, nonlinear, and short duration. So it is very hard to model for these signals with reference patterns. In this paper we separate the whole length of signals into some short duration of constant length with overlapping frame by frame. The 20th LPC(Linear Predictive Coding) coefficients are extracted from the original signals using Durbin algorithm and applied to neural network. The 65% of whole signals were learned and 35% of the signals were tested in the neural network with two hidden layers. The types of the whales for sound classification are Blue whale, Dulsae whale, Gray whale, Humpback whale, Minke whale, and Northern Right whale. Finally, we could obtain more than 83% of classification rate from the test signals.