• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.7
• /
• pp.1387-1393
• /
• 2007

Even if the linearly separable patterns can be classified by the conventional single layer perceptron, the non-linear problems such as XOR can not be classified by it. A fuzzy single layer perceptron can solve the conventional XOR problems by applying fuzzy membership functions. However, in the fuzzy single layer perception, there are a couple disadvantages which are a decision boundary is sometimes vibrating and a convergence may be extremely lowered according to the scopes of the initial values and learning rates. In this paper, for these reasons, we proposed an enhanced fuzzy single layer perceptron algorithm that can prevent from vibration the decision boundary by introducing a bias term and can also reduce the learn time by applying the modified delta rule which include the learning rates and the momentum concept and applying the new linear activation function. Consequently, the simulation results of the XOR and pattern classification problems presented that the proposed method provided the shorter learning time and better convergence than the conventional fuzzy single layer perceptron.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.1
• /
• pp.185-194
• /
• 2014

One in every 10 persons suffer from chronic gastritis in Korea. Endoscopy is most commonly used to diagnose the chronic gastritis. Endoscopic diagnosis is precise but it is accompanied with pain and high cost. According to pulse diagnosis in Traditional East Asian Medicine, health problems in stomach can be diagnosed with radial pulse signals in 'Guan' location in the right wrist, which are non-invasive and cost-effective. In this study, we developed a classification model of chronic gastritis using pulse signals in right 'Guan' location. We used both linear discrimination method and logistic regression model with respect to pulse features obtained with a peak-valley detection algorithm and a Gaussian model. As a result, we obtained sensitivity ranged between 77%~89% and specificity ranged between 72%~83% depending on classification models and feature extraction methods, and the average classification rates were approximately 80%, irrespective of the models. Specifically, the Gaussian model were featured by superior sensitivities (89.1% and 87.5%) while the peak-valley detection method showed superior specificities (82.8% and 81.3%), and the average classification rate (sensitivity + specificity) of the Gaussian model was 80.9% which was 1.2% ahead of the peak-valley method. In conclusion, we obtained a reliable classification model for the chronic gastritis based on the radial pulse feature extraction algorithms, where the Gaussian model was featured by outperformed sensitivity and the peak-valley method was featured by outperformed specificity.

• Journal of Biomedical Engineering Research
• /
• v.23 no.3
• /
• pp.197-205
• /
• 2002

The objectives of this Paper is to implement a diagnostic classifier of differential laryngeal diseases from acoustic signals acquired in a noisy room. For this Purpose, the voice signals of the vowel /a/ were collected from Patients in a soundproof chamber and got mixed with noise. Then, the acoustic Parameters were analyzed, and hierarchical neural networks were applied to the data classification. The classifier had a structure of five-step hierarchical neural networks. The first neural network classified the group into normal and benign or malign laryngeal disease cases. The second network classified the group into normal or benign laryngeal disease cases The following network distinguished polyp. nodule. Palsy from the benign laryngeal cases. Glottic cancer cases were discriminated into T1, T2. T3, T4 by the fourth and fifth networks All the neural networks were based on multilayer perceptron model which classified non-linear Patterns effectively and learned by an error back-propagation algorithm. We chose some acoustic Parameters for classification by investigating the distribution of laryngeal diseases and Pilot classification results of those Parameters derived from MDVP. The classifier was tested by using the chosen parameters to find the optimum ones. Then the networks were improved by including such Pre-Processing steps as linear and z-score transformation. Results showed that 90% of T1, 100% of T2-4 were correctly distinguished. On the other hand. 88.23% of vocal Polyps, 100% of normal cases. vocal nodules. and vocal cord Paralysis were classified from the data collected in a noisy room.

• Geophysics and Geophysical Exploration
• /
• v.13 no.3
• /
• pp.203-218
• /
• 2010

The site categorization and corresponding site amplification factors in the current Korean seismic design guideline are based on provisions for the western United States (US), although the site effects resulting in the amplification of earthquake ground motions are directly dependent on the regional and local site characteristic conditions. In these seismic codes, two amplification factors called site coefficients, $F_a$ and $F_v$, for the short-period band and midperiod band, respectively, are listed according to a criterion, mean shear wave velocity ($V_S$) to a depth of 30 m, into five classes composed of A to E. To suggest a site classification system reflecting Korean site conditions, in this study, systematic site characterization was carried out at four regional areas, Gyeongju, Hongsung, Haemi and Sacheon, to obtain the $V_S$ profiles from surface to bedrock in field and the non-linear soil properties in laboratory. The soil deposits in Korea, which were shallower and stiffer than those in the western US, were examined, and thus the site period in Korea was distributed in the low and narrow band comparing with those in western US. Based on the geotechnical characteristic properties obtained in the field and laboratory, various site-specific seismic response analyses were conducted for total 75 sites by adopting both equivalent-linear and non-linear methods. The analysis results showed that the site coefficients specified in the current Korean provision underestimate the ground motion in the short-period range and overestimate in the mid-period range. These differences can be explained by the differences in the local site characteristics including the depth to bedrock between Korea and western US. Based on the analysis results in this study and the prior research results for the Korean peninsula, new site classification system was developed by introducing the site period as representative criterion and the mean $V_S$ to a depth of shallower than 30 m as additional criterion, to reliably determine the ground motions and the corresponding design spectra taking into account the regional site characteristics in Korea.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.13 no.6
• /
• pp.1333-1342
• /
• 2018

Until now, Electroencephalography(: EEG) has been the most important and convenient method for the diagnosis and treatment of epilepsy. However, it is difficult to identify the wave characteristics of an epileptic EEG signals because it is very weak, non-stationary and has strong background noise. In this paper, we analyse the effect of dimensionality reduction methods on Epileptic EEG feature selection and classification. Three dimensionality reduction methods: Pincipal Component Analysis(: PCA), Kernel Principal Component Analysis(: KPCA) and Linear Discriminant Analysis(: LDA） were investigated. The performance of each method was evaluated by using Support Vector Machine SVM, Logistic Regression(: LR), K-Nearestneighbor(: K-NN), Decision Tree(: DR) and Random Forest(: RF). From the experimental result, PCA recorded 75% of highest accuracy in SVM, LR and K-NN. KPCA recorded 85% of best performance in SVM and K-KNN while LDA achieved 100% accuracy in K-NN. Thus, LDA dimensionality reduction is found to provide the best classification result for epileptic EEG signal.

• Geomechanics and Engineering
• /
• v.30 no.3
• /
• pp.259-272
• /
• 2022

Rockburst is a dynamic, multivariate, and non-linear phenomenon that occurs in underground mining and civil engineering structures. Predicting rockburst is challenging since conventional models are not standardized. Hence, machine learning techniques would improve the prediction accuracies. This study describes decision based uncertainty models to predict rockburst in underground engineering structures using gradient boosting algorithms (GBM). The model input variables were uniaxial compressive strength (UCS), uniaxial tensile strength (UTS), maximum tangential stress (MTS), excavation depth (D), stress ratio (SR), and brittleness coefficient (BC). Several models were trained using different combinations of the input variables and a 3-fold cross-validation resampling procedure. The hyperparameters comprising learning rate, number of boosting iterations, tree depth, and number of minimum observations were tuned to attain the optimum models. The performance of the models was tested using classification accuracy, Cohen's kappa coefficient (k), sensitivity and specificity. The best-performing model showed a classification accuracy, k, sensitivity and specificity values of 98%, 93%, 1.00 and 0.957 respectively by optimizing model ROC metrics. The most and least influential input variables were MTS and BC, respectively. The partial dependence plots revealed the relationship between the changes in the input variables and model predictions. The findings reveal that GBM can be used to anticipate rockburst and guide decisions about support requirements before mining development.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.12
• /
• pp.1853-1858
• /
• 2021

O-rings fill the gaps between mechanical parts. Until now, the sorting of defective products has been performed visually and manually, so classification errors often occur. Therefore, a camera-based defect classification system without human intervention is required. However, a binarization process is required to separate the required region from the background in the camera input image. In this paper, an adaptive binarization technique that considers the surrounding pixel values is applied to solve the problem that single-threshold binarization is difficult to apply due to factors such as changes in ambient lighting or reflections. In addition, the convex hull technique is also applied to compensate for the missing pixel part. And the learning model to be applied to the separated region applies the residual error-based deep learning neural network model, which is advantageous when the defective characteristic is non-linear. It is suggested that the proposed system through experiments can be applied to the automation of O-ring defect detection.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2005.05b
• /
• pp.9-18
• /
• 2005

In order to systematically and visually understand well-known but qualitative and rotatively complicated relationships between synoptic fields in the BAIU season and heavy rainfall events in Japan, these synoptic fields were classified using the Self-Organizing Map (SOM) algorithm. This algorithm can convert complex nonlinear features into simple two-dimensional relationships, and was followed by the application of the clustering techniques of the U-matrix and the K-means. It was assumed that the meteorological field patterns be simply expressed by the spatial distribution of wind components at the 850 hPa level and Precipitable Water (PW) in the southwestern area including Kyushu in Japan. Consequently, the synoptic fields could be divided into eight kinds of patterns (clusters). One of the clusters has the notable spatial feature represented by high PW accompanied by strong wind components known as Low-Level Jet (LLJ). The features of this cluster indicate a typical meteorological field pattern that frequently causes disastrous heavy rainfall in Kyushu in the rainy season. From these results, the SOM technique may be an effective tool for the classification of complicated non-linear synoptic fields.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.35 no.1
• /
• pp.57-65
• /
• 2012

The relationship between product quality/function and customer satisfaction has been considered an important point for the new product development. The seminal paper by Kano was the first to thoroughly address the non-linear relationship between product performance/function and customer satisfaction. In the analysis framework of the original Kano model, five factors are assumed, among which indifference factor occupies 40% in the classification scheme. When we analyze survey response using Kano model, many quality attributes can be resulted in indifference factor. This implies that some attributes which are meaningful tend to be classified as indifferent attributes for the customer satisfaction. In order to tackle this problem, a modified Kano model is proposed by reducing the indifference factor. The modified Kano model can be robust for the survey response. A survey is performed for the quality attributes of the smart phone. The response is analyzed and compared based on the original and modified Kano model. The surveyed quality characteristics of the smart phone are performance related attributes, application programs, functional attributes and subjective emotional quality attributes. Many quality attributes classified as indifference factor in the original model are classified as attractive, must-be, and expected factors, respectively.

• Journal of the Ergonomics Society of Korea
• /
• v.24 no.4
• /
• pp.31-37
• /
• 2005

Resulting anthropometric data recently measured and cataloged through 5th national anthropometric survey that is called Size Korea is highly useful in clothing industries. This study aims at suggesting a statistical methodology for apparel sizing that reflects recent changes in Korean anthropometry and improves customer fitness. Based on previous research on body types such as triangular, rectangular, inverted body types, etc., factors that represent human sizes were extracted and then clustered into groups by their body types. These body type-based groups with respect to the factors obtained yielded a sizing system of which the interval of each factor is of equi-distance by their factor scores. However, each interval of the sizing system is non-linear in terms of individual anthropometric variables. The sizing system being proposed in this study was compared to that of KS K 0050 and had a broader coverage for the Korean population surveyed. The apparel sizing scheme is expected to improve customer fitness when applied to garment sizing and to provide more information on what percentage of population is included in each classification.