• The Journal of Korean Institute of Next Generation Computing
• /
• v.15 no.4
• /
• pp.7-17
• /
• 2019

Machine learning is one of the most accurate techniques to predict and analyze various phenomena. K-means clustering is a kind of machine learning technique that classifies given data into clusters of similar data. Because it is desirable to perform an analysis based on a lot of data for better performance, K-means clustering can be performed in a model with a server that calculates the centroids of the clusters, and a number of clients that provide data to server. However, this model has the problem that if the clients' data are associated with private information, the server can infringe clients' privacy. In this paper, to solve this problem in a model with a number of clients, we propose a privacy-preserving K-means clustering method that can perform machine learning, concealing private information using homomorphic encryption.

• Journal of Information Processing Systems
• /
• v.16 no.6
• /
• pp.1424-1436
• /
• 2020

Although huge progress has been made in current image segmentation work, there are still no efficient segmentation strategies for tree image which is taken from natural environment and contains complex background. To improve those problems, we propose a method for tree image segmentation combining adaptive mean shifting with image abstraction. Our approach perform better than others because it focuses mainly on the background of image and characteristics of the tree itself. First, we abstract the original tree image using bilateral filtering and image pyramid from multiple perspectives, which can reduce the influence of the background and tree canopy gaps on clustering. Spatial location and gray scale features are obtained by step detection and the insertion rule method, respectively. Bandwidths calculated by spatial location and gray scale features are then used to determine the size of the Gaussian kernel function and in the mean shift clustering. Furthermore, the flood fill method is employed to fill the results of clustering and highlight the region of interest. To prove the effectiveness of tree image abstractions on image clustering, we compared different abstraction levels and achieved the optimal clustering results. For our algorithm, the average segmentation accuracy (SA), over-segmentation rate (OR), and under-segmentation rate (UR) of the crown are 91.21%, 3.54%, and 9.85%, respectively. The average values of the trunk are 92.78%, 8.16%, and 7.93%, respectively. Comparing the results of our method experimentally with other popular tree image segmentation methods, our segmentation method get rid of human interaction and shows higher SA. Meanwhile, this work shows a promising application prospect on visual reconstruction and factors measurement of tree.

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.6
• /
• pp.566-571
• /
• 2009

This paper proposes noise-robust fast speaker adaptation method based on the eigenvoice framework in various noisy environments. The proposed method is focused on de-noising and environment clustering. Since the de-noised adaptation DB still has residual noise in itself, environment clustering divides the noisy adaptation data into similar environments by a clustering method using the cepstral mean of non-speech segments as a feature vector. Then each adaptation data in the same cluster is used to build an environment-clustered speaker adapted (SA) model. After selecting multiple environmentally clustered SA models which are similar to test environment, the speaker adaptation based on an appropriate linear combination of clustered SA models is conducted. According to our experiments, we observe that the proposed method provides error rate reduction of $40{\sim}59%$ over baseline with speaker independent model.

• Journal of KIISE:Software and Applications
• /
• v.27 no.5
• /
• pp.529-542
• /
• 2000

Recent years, the use of multimedia information is rapidly increasing, and the video media is the most rising one than any others, and this field Integrates all the media into a single data stream. Though the availability of digital video is raised largely, it is very difficult for users to make the effective video access, due to its length and unstructured video format. Thus, the minimal interaction of users and the explicit definition of video structure is a key requirement in the lately developing image and video management systems. This paper defines the terms and hierarchical video structure, and presents the system, which construct the clustering-based video hierarchy, which facilitate users by browsing the summary and do a random access to the video content. Instead of using a single feature and domain-specific thresholds, we use multiple features that have complementary relationship for each other and clustering-based methods that use normalization so as to interact with users minimally. The stage of shot boundary detection extracts multiple features, performs the adaptive filtering process for each features to enhance the performance by eliminating the false factors, and does k-means clustering with two classes. The shot list of a result after the proposed procedure is represented as the video hierarchy by the intelligent unsupervised clustering technique. We experimented the static and the dynamic movie videos that represent characteristics of various video types. In the result of shot boundary detection, we had almost more than 95% good performance, and had also rood result in the video hierarchy.

• Journal of the Korea Society for Simulation
• /
• v.29 no.1
• /
• pp.47-55
• /
• 2020

This paper presents a real-time autonomous computation of shot numbers and aiming points against multiple soft targets on grounds by applying an unsupervised learning, k-mean clustering and Monte carlo simulation. For this computation, a 100 × 200 square meters size of virtual battlefield is created where an augmented enemy infantry platoon unit attacks, defences, and is scatted, and a virtual weapon with a lethal range of 15m is modeled. In order to determine damage types of the enemy unit: no damage, light wound, heavy wound and death, Monte carlo simulation is performed to apply the Carlton damage function for the damage effect of the soft targets. In addition, in order to achieve the damage effectiveness of the enemy units in line with the commander's intention, the optimal shot numbers and aiming point locations are calculated in less than 0.4 seconds by applying the k-mean clustering and repetitive Monte carlo simulation. It is hoped that this study will help to develop a system that reduces the decision time for 'detection-decision-shoot' process in battalion-scaled combat units operating Dronebot combat system.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.1
• /
• pp.272-287
• /
• 2016

Recently, various and effective contrast based salient object detection models to focus on a single target have been proposed. However, there is a lack of research on detection of multiple objects, and also it is a more challenging task than single target process. In the multiple target problem, we are confronted by new difficulties caused by distinct difference between properties of objects. The characteristic of existing models depending on the global maximum distribution of data point would become a drawback for detection of multiple objects. In this paper, by analyzing limitations of the existing methods, we have devised three main processes to detect multiple salient objects. In the first stage, regional features are extracted from over-segmented regions. In the second stage, the regional features are categorized into homogeneous cluster using the mean-shift algorithm with the kernel function having various sizes. In the final stage, we compute saliency scores of the categorized regions using only spatial features without the contrast features, and then all scores are integrated for the final salient regions. In the experimental results, the scheme achieved superior detection accuracy for the SED2 and MSRA-ASD benchmarks with both a higher precision and better recall than state-of-the-art approaches. Especially, given multiple objects having different properties, our model significantly outperforms all existing models.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.18 no.1
• /
• pp.59-69
• /
• 2018

In this paper, we present a robust beamforming design to tackle the weighted sum-rate maximization (WSRM) problem in a multicell multiple-input multiple-output (MIMO) - non-orthogonal multipleaccess (NOMA) downlink system for 5G wireless communications. This work consider the imperfectchannel state information (CSI) at the base station (BS) by adding uncertainties to channel estimation matrices as the worst-case model i.e., singular value uncertainty model (SVUM). With this observation, the WSRM problem is formulated subject to the transmit power constraints at the BS. The objective problem is known as on-deterministic polynomial (NP) problem which is difficult to solve. We propose an robust beam forming design which establishes on majorization minimization (MM) technique to find the optimal transmit beam forming matrix, as well as efficiently solve the objective problem. In addition, we also propose a joint user clustering and power allocation (JUCPA) algorithm in which the best user pair is selected as a cluster to attain a higher sum-rate. Extensive numerical results are provided to show that the proposed robust beamforming design together with the proposed JUCPA algorithm significantly increases the performance in term of sum-rate as compared with the existing NOMA schemes and the conventional orthogonal multiple access (OMA) scheme.

• Journal of Computing Science and Engineering
• /
• v.2 no.2
• /
• pp.109-136
• /
• 2008

In this survey paper, we discuss biomedical ontologies and major text mining techniques applied to biomedicine and healthcare. Biomedical ontologies such as UMLS are currently being adopted in text mining approaches because they provide domain knowledge for text mining approaches. In addition, biomedical ontologies enable us to resolve many linguistic problems when text mining approaches handle biomedical literature. As the first example of text mining, document clustering is surveyed. Because a document set is normally multiple topic, text mining approaches use document clustering as a preprocessing step to group similar documents. Additionally, document clustering is able to inform the biomedical literature searches required for the practice of evidence-based medicine. We introduce Swanson's UnDiscovered Public Knowledge (UDPK) model to generate biomedical hypotheses from biomedical literature such as MEDLINE by discovering novel connections among logically-related biomedical concepts. Another important area of text mining is document classification. Document classification is a valuable tool for biomedical tasks that involve large amounts of text. We survey well-known classification techniques in biomedicine. As the last example of text mining in biomedicine and healthcare, we survey information extraction. Information extraction is the process of scanning text for information relevant to some interest, including extracting entities, relations, and events. We also address techniques and issues of evaluating text mining applications in biomedicine and healthcare.

• Korean Journal of Health Education and Promotion
• /
• v.33 no.1
• /
• pp.1-12
• /
• 2016

Objectives: This study was to examine by gender the clustering patterns and correlates of healthy lifestyle clusters and the relationships between healthy lifestyle clusters and depressive symptoms in middle-aged and older adults. Methods: The observed/expected ratio of physical activity, smoking, and alcohol consumption were calculated to analyze clustering effects. The correlates of those healthy lifestyle clusters were evaluated using logistic regression models, and the relationship between those healthy lifestyle clusters and depressive symptoms was investigated using multiple regressions by gender. Results: Based on the guidelines this study adopted, we obtained three healthy lifestyle clusters: active healthy lifestyle; passive healthy lifestyle; and unhealthy lifestyle. All three clusters were found in men, but two in women, who did not have an unhealthy lifestyle cluster. High socio-economic status was positively related to healthy lifestyle clusters. Social participation and residence location (in men) and marital status (in women) were significant factors. Having an active or a passive healthy lifestyle was negatively associated with depressive symptoms in women, but such a relationship was not observed in men. Conclusions: The study findings imply that health promotion programs for middle-aged and older adults in Korea should be comprehensive and integrated, considering healthy lifestyle clusters and gender differences.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.7
• /
• pp.1150-1156
• /
• 2008

This paper introduces the PC clustering of the SIMD structure for a distributed processing of on-line contingency to assess a static security of a power system. To execute on-line contingency analysis of a large-scale power system, we need to use high-speed execution device. Therefore, we constructed PC-cluster system using PC clustering method of the SIMD structure and applied to a power system, which relatively shows high quality on the high-speed execution and has a low price. SIMD(single instruction stream, multiple data stream) is a structure that processes are controlled by one signal. The PC cluster system is consisting of 8 PCs. Each PC employs the 2 GHz Pentium 4 CPU and is connected with the others through ethernet switch based fast ethernet. Also, we consider N-1 line contingency that have high potentiality of occurrence realistically. We propose the distributed process algorithm of the SIMD structure for reducing too much execution time on the on-line N-1 line contingency analysis in the large-scale power system. And we have verified a usefulness of the proposed algorithm and the constructed PC cluster system through IEEE 39 and 118 bus system.