• Wind and Structures
• /
• v.36 no.6
• /
• pp.393-404
• /
• 2023

Skewness and kurtosis are important higher-order statistics for simulating non-Gaussian wind pressure series on low-rise buildings, but their predictions are less studied in comparison with those of the low order statistics as mean and rms. The distribution gradients of skewness and kurtosis on roofs are evidently higher than those of mean and rms, which increases their prediction difficulty. The conventional artificial neural networks (ANNs) used for predicting mean and rms show unsatisfactory accuracy in predicting skewness and kurtosis owing to the limited capacity of shallow learning of ANNs. In this work, the deep neural networks (DNNs) model with the ability of deep learning is introduced to predict the skewness and kurtosis on a low-rise building. For obtaining the optimal generalization of the DNNs model, the hyper parameters are automatically determined by Bayesian Optimization (BO). Moreover, for providing a benchmark for future studies on predicting higher order statistics, the data sets for training and testing the DNNs model are extracted from the internationally open NIST-UWO database, and the prediction errors of all taps are comprehensively quantified by various error metrices. The results show that the prediction accuracy in this study is apparently better than that in the literature, since the correlation coefficient between the predicted and experimental results is 0.99 and 0.75 in this paper and the literature respectively. In the untrained cornering wind direction, the distributions of skewness and kurtosis are well captured by DNNs on the whole building including the roof corner with strong non-normality, and the correlation coefficients between the predicted and experimental results are 0.99 and 0.95 for skewness and kurtosis respectively.

• Journal of the Korean GEO-environmental Society
• /
• v.25 no.11
• /
• pp.5-12
• /
• 2024

In this study, the liquefaction potential index was assessed using actual borehole data and seismic waves, and a predictive model was developed based on machine learning methods. A total of 10 features were selected including factors reflecting the characteristics of the seismic waves. To identify candidate methods, a preliminary test was conducted using commonly used machine learning methods for regression, followed by Bayesian optimization to optimize the hyperparameters for these candidate methods. Among artificial neural networks, Gaussian process regression, and random forest, it was found that the random forest effectively predicted the liquefaction potential index, as indicated by a low root mean square error, a high coefficient of determination, and considerations regarding overfitting. However, it was noted that the model tends to underestimate the liquefaction potential index when the index was 5 or higher.

• Journal of the Korean Statistical Society
• /
• v.30 no.2
• /
• pp.179-192
• /
• 2001

When a statistical model has a hierarchical structure such as multilayer perceptrons in neural networks or Gaussian mixture density representation, the model includes distribution with unidentifiable parameters when the structure becomes redundant. Since the exact structure is unknown, we need to carry out statistical estimation or learning of parameters in such a model. From the geometrical point of view, distributions specified by unidentifiable parameters become a singular point in the parameter space. The problem has been remarked in many statistical models, and strange behaviors of the likelihood ratio statistics, when the null hypothesis is at a singular point, have been analyzed so far. The present paper studies asymptotic behaviors of the maximum likelihood estimator and the Bayesian predictive estimator, by using a simple cone model, and show that they are completely different from regular statistical models where the Cramer-Rao paradigm holds. At singularities, the Fisher information metric degenerates, implying that the cramer-Rao paradigm does no more hold, and that he classical model selection theory such as AIC and MDL cannot be applied. This paper is a first step to establish a new theory for analyzing the accuracy of estimation or learning at around singularities.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.7
• /
• pp.3076-3092
• /
• 2020

An IKPCA-ELM-based intrusion detection method is developed to address the problem of the low accuracy and slow speed of intrusion detection caused by redundancies and high dimensions of data in the network. First, in order to reduce the effects of uneven sample distribution and sample attribute differences on the extraction of KPCA features, the sample attribute mean and mean square error are introduced into the Gaussian radial basis function and polynomial kernel function respectively, and the two improved kernel functions are combined to construct a hybrid kernel function. Second, an improved particle swarm optimization (IPSO) algorithm is proposed to determine the optimal hybrid kernel function for improved kernel principal component analysis (IKPCA). Finally, IKPCA is conducted to complete feature extraction, and an extreme learning machine (ELM) is applied to classify common attack type detection. The experimental results demonstrate the effectiveness of the constructed hybrid kernel function. Compared with other intrusion detection methods, IKPCA-ELM not only ensures high accuracy rates, but also reduces the detection time and false alarm rate, especially reducing the false alarm rate of small sample attacks.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.6
• /
• pp.3055-3073
• /
• 2019

In this paper, a new user identification method is presented using real environmental human-computer-interaction (HCI) behavior data to improve method usability. User behavior data in this paper are collected continuously without setting experimental scenes such as text length, action number, etc. To illustrate the characteristics of real environmental HCI data, probability density distribution and performance of keyboard and mouse data are analyzed through the random sampling method and Support Vector Machine(SVM) algorithm. Based on the analysis of HCI behavior data in a real environment, the Multiple Kernel Learning (MKL) method is first used for user HCI behavior identification due to the heterogeneity of keyboard and mouse data. All possible kernel methods are compared to determine the MKL algorithm's parameters to ensure the robustness of the algorithm. Data analysis results show that keyboard data have a narrower range of probability density distribution than mouse data. Keyboard data have better performance with a 1-min time window, while that of mouse data is achieved with a 10-min time window. Finally, experiments using the MKL algorithm with three global polynomial kernels and ten local Gaussian kernels achieve a user identification accuracy of 83.03% in a real environmental HCI dataset, which demonstrates that the proposed method achieves an encouraging performance.

• Journal of Multimedia Information System
• /
• v.6 no.2
• /
• pp.49-60
• /
• 2019

Plants are very crucial for life on Earth. There is a wide variety of plant species available, and the number is increasing every year. Species knowledge is a necessity of various groups of society like foresters, farmers, environmentalists, educators for different work areas. This makes species identification an interdisciplinary interest. This, however, requires expert knowledge and becomes a tedious and challenging task for the non-experts who have very little or no knowledge of the typical botanical terms. However, the advancements in the fields of machine learning and computer vision can help make this task comparatively easier. There is still not a system so developed that can identify all the plant species, but some efforts have been made. In this study, we also have made such an attempt. Plant identification usually involves four steps, i.e. image acquisition, pre-processing, feature extraction, and classification. In this study, images from Swedish leaf dataset have been used, which contains 1,125 images of 15 different species. This is followed by pre-processing using Gaussian filtering mechanism and then texture and color features have been extracted. Finally, classification has been done using Multiclass-support vector machine, which achieved accuracy of nearly 93.26%, which we aim to enhance further.

• The Bulletin of The Korean Astronomical Society
• /
• v.46 no.2
• /
• pp.74.2-75
• /
• 2021

We demonstrate that a deep learning classifier that only uses to gravitational wave (GW) detectors auxiliary channel data can distinguish various types of non-Gaussian noise transients (glitches) with significant accuracy, i.e., ≳ 80%. The classifier is implemented using the multi-scale neural networks (MSNN) with PyTorch. The glitches appearing in the GW strain data have been one of the main obstacles that degrade the sensitivity of the gravitational detectors, consequently hindering the detection and parameterization of the GW signals. Numerous efforts have been devoted to tracking down their origins and to mitigating them. However, there remain many glitches of which origins are not unveiled. We apply the MSNN classifier to the auxiliary channel data corresponding to publicly available GravitySpy glitch samples of LIGO O1 run without using GW strain data. Investigation of the auxiliary channel data of the segments that coincide to the glitches in the GW strain channel is particularly useful for finding the noise sources, because they record physical and environmental conditions and the status of each part of the detector. By only using the auxiliary channel data, this classifier can provide us with the independent view on the data quality and potentially gives us hints to the origins of the glitches, when using the explainable AI technique such as Layer-wise Relevance Propagation or GradCAM.

• Journal of Platform Technology
• /
• v.11 no.5
• /
• pp.110-125
• /
• 2023

In this paper, we utilize time-series aircraft location data measured based on 2019 domestic airports to analyze Go-Around and UOC_D situations during the approach phase of domestic airports. Various clustering-based machine learning techniques are applied to determine the most appropriate analysis method for domestic aviation data through experimentation. The ADS-B sensor is solely employed to measure aircraft positions. We designed a model using clustering algorithms such as K-Means, GMM, and DBSCAN to classify abnormal situations. Among them, the RF model showed the best performance overseas, but through experiments, it was confirmed that the GMM showed the highest classification performance for domestic aviation data by reflecting the aspects specialized in domestic terrain.

• Proceedings of the KIEE Conference
• /
• 2004.11c
• /
• pp.382-384
• /
• 2004

In this parer, we propose a clustering that perform algorithm using new convergence properties. For detection and optimization of cluster, we use to similarity measure with cumulative probability and to inference the its parameters with MLE. A merits of using the cumulative probability in our method is very effectiveness that robust to noise or unnecessary data for inference the parameters. And we adopt similarity threshold to converge the number of cluster that is enable to past convergence and delete the other influence for this learning algorithm. In the simulation, we show effectiveness of our algorithm for convergence and optimization of cluster in riven data set.

• Industrial Engineering and Management Systems
• /
• v.15 no.2
• /
• pp.148-155
• /
• 2016

Existing power plants may consume significant amounts of fuel and require high operating costs, partly because of poor electrical power output estimates. This paper suggests a continuous conditional random field (C-CRF) model to predict more precisely the full-load electrical power output of a base load operated combined cycle power plant. We introduce three feature functions to model association potential and one feature function to model interaction potential. Together, these functions compose the C-CRF model, and the model is transformed into a multivariate Gaussian distribution with which the operation parameters can be modeled more efficiently. The performance of our model in estimating power output was evaluated by means of a real dataset and our model outperformed existing methods. Moreover, our model can be used to estimate confidence intervals of the predicted output and calculate several probabilities.