• The Journal of the Acoustical Society of Korea
• /
• v.35 no.4
• /
• pp.303-309
• /
• 2016

This paper proposes a novel approach to the problem of audio signal clustering using a stacked autoencoder. The proposed stacked autoencoder learns an efficient representation for the input signal, enables clustering constituent signals with similar characteristics, and therefore the original sources can be separated based on the clustering results. STFT (Short-Time Fourier Transform) is performed to extract time-frequency spectrum, and rectangular windows at all the possible locations are used as input values to the autoencoder. The outputs at the middle, encoding layer, are used to cluster the rectangular windows and the original sources are separated by the Wiener filters derived from the clustering results. Source separation experiments were carried out in comparison to the conventional NMF (Non-negative Matrix Factorization), and the estimated sources by the proposed method well represent the characteristics of the orignal sources as shown in the time-frequency representation.

• Journal of Broadcast Engineering
• /
• v.25 no.5
• /
• pp.734-741
• /
• 2020

This article proposes a new pedestrian avoidance algorithm for social robots that coexist and communicate with humans and do not induce stress caused by invasion of psychological safety distance(Social Distance). To redefine the pedestrian model, pedestrians are clustered according to the pedestrian's gait characteristics(straightness, speed) and a social distance is defined for each pedestrian cluster. After modeling pedestrians(obstacles) with the social distances, integrated navigation algorithm is completed by applying the newly defined pedestrian model to commercial obstacle avoidance and path planning algorithms. To show the effectiveness of the proposed algorithm, two commercial obstacle avoidance & path planning algorithms(the Dynamic Window Approach (DWA) algorithm and the Timed Elastic Bands (TEB) algorithm) are used. Four cases were experimented in applying and non-applying the new pedestrian model, respectively. Simulation results show that the proposed algorithm can significantly reduce the stress index of pedestrians without loss of traveling time.

• The KIPS Transactions:PartC
• /
• v.11C no.7 s.96
• /
• pp.923-930
• /
• 2004

Wireless sensor networks is a core technology of ubiquitous computing which enables the network to aware the different kind of context by integrating exiting wired/wireless infranet with various sensor devices and connecting collected environmental data with applications. However it needs an energy-efficient approach in network layer to maintain the dynamic ad hoc network and to maximize the network lifetime by using energy constrained node. Cluster-based data aggregation and routing are energy-efficient solution judging from architecture of sensor networks and characteristics of data. In this paper. we propose a new distributed clustering algorithm in using distance from the sink. This algorithm shows that it can balance energy dissipation among nodes while minimizing the overhead. We verify that our clustering is more en-ergy-efficient and thus prolongs the network lifetime in comparing our proposed clustering to existing probabilistic clustering for sensor network via simulation.

• Journal of KIISE:Databases
• /
• v.32 no.5
• /
• pp.498-508
• /
• 2005

A shape data is used one oi most important feature for image retrieval as data to reflect meaning of image. Especially, structural feature of shape is widely studied because it represents primitive properties of shape and relation information between basic units well. However, most structural features of shape have the problem that it is not able to guarantee an efficient search time because the features are expressed as graph or tree. In order to solve this problem, we generate curvature-based morphological graph, End design key to cluster shapes from this graph. Proposed this graph have contour features and morphological features of a shape. Shape retrieval is accomplished by stages. We reduce a search space through clustering, and determine total similarity value through pattern matching of external curvature. Various experiments show that our approach reduces computational complexity and retrieval cost.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.2
• /
• pp.171-177
• /
• 2011

Accuracy and computing time are considerable issues in machine learning. In general, the computing time for data analysis is increased in proportion to the size of given data. So, we need a sampling approach to reduce the size of training data. But, the accuracy of constructed model is decreased by going down the data size simultaneously. To solve this problem, we propose a new statistical sampling method having similar performance to the total data. We suggest a rule to select optimal sampling techniques according to given data structure. This paper shows a sampling method for reducing computing time with keeping the most of accuracy using cluster sampling, stratified sampling, and systematic sampling. We verify improved performance of proposed method by accuracy and computing time between sample data and total data using objective machine learning data sets.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.4
• /
• pp.568-573
• /
• 2009

Marine accidents have always caused huge economic losses, as well as environmental pollution. Prevention of marine accidents has become a focus of argumentation. The analysis of past accident cases, reviewing the experience and lessons, is important and necessary for preventing marine accidents. With the same subject above, the Korean Maritime Safety Tribunal provides for past marine accidents' written judgments and analysis of judgment and associated retrieval system on its homepage. In these systems, the name of the ship, accident occurrence time, accident pattern or related keywords are used as search conditions. However, most of the marine events' happening were not due to a single reason, but multiple ones. In addition, one marine event could often come under several categories. In this case, now the retrieval systems' DB is used on the Korean Maritime Safety Tribunal homepage was built based on single category and failed to be able to retrieve according to multiple reasons or multiple categories. In order to solve this problem, a more practical retrieval approach might be needed. Therefore, in this paper, a new retrieval system will be proposed, which using the linguistic label to describe the cluster after analyzing the relational properties between marine accidents and clustering by FCM algorithm, and then adding an interface to allow users to get the results they want through choosing multiple reasons or multiple categories.

• Journal of Energy Engineering
• /
• v.16 no.3
• /
• pp.149-154
• /
• 2007

An experimental study on the air-side pressure drop of slit fin-tube heat exchanger has been carried out. The data reduction methodology for air-side pressure drop in the literature is not based on a consistent approach. This paper focuses on method of data reduction to obtain the air-side performance of fin-tube heat exchanger using R22 and recommends standard procedures for dry and wet surface pressure drop estimation in fin-tube heat exchanger. A comparison was made between the predictions of previously proposed empirical correlations and experimental data for the air-side pressure drop on design conditions of condenser and evaporator. Results are pre-sented as plots of friction f-factor against Reynolds number based on the fin collar outside diameter and compared with previous studies. The data covers a range of refrigerant mass fluxes of $150{\sim}250\;kg/m^2s$ with air flows at velocity ranges from 0.38 m/s to 1.6 m/s.

• Korean Journal of Acupuncture
• /
• v.34 no.2
• /
• pp.88-99
• /
• 2017

Objectives : Saam acupuncture initiated by Saamdoin in $17^{th}$ century is one of the most widely adopted acupuncture techniques used by Korean medical doctors in clinic. Our study aimed to analyze the application of the Saam acupuncture method to pain diseases based on the literature data. Methods : Based on the contents described in "(Do Hae Kyo Kam) Saam's acupuncture method", the texts related to pain condition were analyzed. The frequency of prescription of Saam acupuncture method was analyzed, and then the relationships between each acupoint were visualized by network analysis and hierarchical cluster analysis for the quantitative aspect. Results and conclusions: In our study, Lung tonifying and Liver tonifying acupuncture were the most frequently used method for the treatment of pain disease. As the acupoints, BL66 and SI5 were used the most frequently. It was found that visceral pattern identification was considered as the most important factor in the selection of the Saam acupuncture method. Network analysis and hierarchical clustering analysis showed that each acupoint was closely related to other acupoints, and most of them were connected more closely according to the method of Saam acupuncture operation. The experiential prescriptions of Saam acupuncture were classified as an independent group. In the future, fundamental research on the principle of Saam acupuncture method is needed for the various diseases, and research for the clinical efficacy and the mechanism of Saam acupuncture method should be preceded.

• Analytical Science and Technology
• /
• v.33 no.2
• /
• pp.98-107
• /
• 2020

Methamphetamine (MA) is currently the most abused illicit drug in Korea. MA is produced by chemical synthesis, and the final target drug that is produced contains small amounts of the precursor chemicals, intermediates, and by-products. To identify and quantify these trace compounds in MA seizures, a practical and feasible approach for conducting chromatographic fingerprinting with a suite of traditional chemometric methods and recently introduced machine learning approaches was examined. This was achieved using gas chromatography (GC) coupled with a flame ionization detector (FID) and mass spectrometry (MS). Following appropriate examination of all the peaks in 71 samples, 166 impurities were selected as the characteristic components. Unsupervised (principal component analysis (PCA), hierarchical cluster analysis (HCA), and K-means clustering) and supervised (partial least squares-discriminant analysis (PLS-DA), orthogonal partial least squares-discriminant analysis (OPLS-DA), support vector machines (SVM), and deep neural network (DNN) with Keras) chemometric techniques were employed for classifying the 71 MA seizures. The results of the PCA, HCA, K-means clustering, PLS-DA, OPLS-DA, SVM, and DNN methods for quality evaluation were in good agreement. However, the tested MA seizures possessed distinct features, such as chirality, cutting agents, and boiling points. The study indicated that the established qualitative and semi-quantitative methods will be practical and useful analytical tools for characterizing trace compounds in illicit MA seizures. Moreover, they will provide a statistical basis for identifying the synthesis route, sources of supply, trafficking routes, and connections between seizures, which will support drug law enforcement agencies in their effort to eliminate organized MA crime.

• The Journal of the Korea Contents Association
• /
• v.12 no.2
• /
• pp.19-29
• /
• 2012

Currently, cloud computing technology is being supplied in various service forms and it is becoming a ground breaking service which provides usage of storage service, data and software while user is not involved in technical background such as physical location of service or system environment. cloud computing technology has advantages that it can use easily as many IT resources as it wants freely regardless of hardware issues required by a variety of systems and service level required by infrastructure. Also, since it has a strength that it can choose usage of resource about business model due to various internet-based technologies, provisioning technology and virtualization technology are being paid attention as main technologies. These technologies are ones of important technology elements which help web-based users approach freely and install according to user environment. Therefore, this thesis introduces software-related technologies and architectures in an aspect of grid for building up high availability cloud computing environment by analysis about cloud computing technology trend.