• The KIPS Transactions:PartD
• /
• v.19D no.2
• /
• pp.151-160
• /
• 2012

Most of the distributed high-dimensional indexing structures provide a reasonable search performance especially when the dataset is uniformly distributed. However, in case when the dataset is clustered or skewed, the search performances gradually degrade as compared with the uniformly distributed dataset. We propose a method of improving the k-nearest neighbor search performance for the distributed vector approximation-tree based on the strongly clustered or skewed dataset. The basic idea is to compute volumes of the leaf nodes on the top-tree of a distributed vector approximation-tree and to assign different number of bits to them in order to assure an identification performance of vector approximation. In other words, it can be done by assigning more bits to the high-density clusters. We conducted experiments to compare the search performance with the distributed hybrid spill-tree and distributed vector approximation-tree by using the synthetic and real data sets. The experimental results show that our proposed scheme provides consistent results with significant performance improvements of the distributed vector approximation-tree for strongly clustered or skewed datasets.

• Journal of KIISE:Information Networking
• /
• v.36 no.5
• /
• pp.437-455
• /
• 2009

In the near future, various multicast based services will be provided over clustered wireless networks. To provide multicast services in secure manner, multicast messages are encrypted by using group key which is shared by group members. Therefore, various group key management schemes have been introduced until now. Among them, tree based approach is one of the most representative paradigms in group key management. Traditional tree based approaches effectively reduce rekeying message transmissions of the key distribution center. However, they do not consider the network bandwidth used for transmitting the rekeying messages. In this paper, we firstly present formulas that describe bandwidth consumption of tree based group key management scheme. Based on our formulations, we propose a bandwidth efficient key tree management scheme for clustered wireless networks where membership changes occur frequently. Simulation results show that our scheme effectively reduces the bandwidth consumption used for rekeying compared to existing key tree schemes.

• Korean Journal of Veterinary Research
• /
• v.43 no.2
• /
• pp.219-230
• /
• 2003

Porcine epidemic diarrhea virus(PED), a member of Coronaviridea, is the etiological agent of enteropathogenic diarrhea in swine. The purpose of this study was to investigate genetic characteristic of PEDV isolated in Korea. Nucleocapsid(N) gene and membrane (M) gene of recent Korean PEDV strains isolated in 2001 were amplified, cloned, sequenced and analyzed. N gene of seven Korean PEDV field isolates bad 94.5% to 99.4% nucleotide and 92.4% to 99.4% amino acid sequence homology each other. Nucleotide and amino acid sequences of Korean field PEDVs were different from published foreign PEDVs, showing 95.1% to 98.0% nucleotide and 93.5% to 97.6% amino acid sequence homology. By phylogenetic tree analysis on based nucleotide sequences, PEDVs were clustered into four groups. By phylogenetic tree analysis based on amino acid sequences. PEDVs were clustered into five groups. M gene of our Korean PEDV field isolates had 99.6% to 100% nucleotide and 98.7% to 100% amino acid sequence homology each other. Nuclotide and amino acid sequences of Korean field PEDVs were different from published foreign PEDVs, showing 98.5% to 98.8% nucleotide and 97.3% to 97.8% amino acid sequence homology. By phylogenetic tree analysis based on nucleotide and amino acid sequences, PEDVs were clustered into two groups which were Korean PEDV isolate group and foreign PEDV isolate group.

• Animal cells and systems
• /
• v.4 no.1
• /
• pp.31-37
• /
• 2000

The genetic variations among six species of Rana from Korea (R. nigro-maculata, R. piancyi, R. dybowskii, R. sp, R. rugosa type A, B and R. amurensis) were investigated using 499 bases of mitochondrial DNA sequences for ND2 (NADH dehydrogenase subunit 2) gene and $tRNA^{Trp}$ gene. Partial sequences of ND2 gene (427 bp) and full sequences of $tRNA^{Trp}$ gene (73 bp) were identified. The level of sequence divergences ranged from 0.2 to 5.2% within species and 4.9-28.0% among 6 species of the genus Rana. The $tRNA^{Trp}$ gene of the genus Rana was composed of 77 nucleotides which showed a two dimensional "cloverleaf" structure. The secondary structure of $tRNA^{Trp}$ was not found compensatory changes which could potentially confound phylogenetic inference. In the neighborjoining tree, brown frogs were clustered first with the level of sequence divergence of 13.20% between R. amurensis and R. dybowskii, and 9% between R. dybowskii and R. sp. supported by 99% bootstrap iterations, respectively. R. nigromaculata and R. plancyi were clustered into another group with 5.1% divergence supported by 100% bootstrap iteration. R. rugosa A 8nd B types were grouped by 4.9% divergence and clustered into the last group with other two groups with 100% bootstrap iterations.

• Speech Sciences
• /
• v.10 no.1
• /
• pp.71-84
• /
• 2003

In the acoustic modeling for large vocabulary speech recognition, a sparse data problem caused by a huge number of context-dependent (CD) models usually leads the estimated models to being unreliable. In this paper, we develop a new clustering method based on the C45 decision-tree learning algorithm that effectively encapsulates the CD modeling. The proposed scheme essentially constructs a supervised decision rule and applies over the pre-clustered triphones using the C45 algorithm, which is known to effectively search through the attributes of the training instances and extract the attribute that best separates the given examples. In particular, the data driven method is used as a clustering algorithm while its result is used as the learning target of the C45 algorithm. This scheme has been shown to be effective particularly over the database of low unknown-context ratio in terms of recognition performance. For speaker-independent, task-independent continuous speech recognition task, the proposed method reduced the percent accuracy WER by 3.93% compared to the existing rule-based methods.

• Journal of KIISE:Databases
• /
• v.31 no.1
• /
• pp.13-22
• /
• 2004

With the proliferation of multimedia data, there is an increasing need to support the indexing and retrieval of high-dimensional image data. Although there have been many efforts, the performance of existing multidimensional indexing methods is not satisfactory in high dimensions. Thus the dimensionality reduction and the approximate solution methods were tried to deal with the so-called dimensionality curse. But these methods are inevitably accompanied by the loss of precision of query results. Therefore, recently, the vector approximation-based methods such as the VA- file and the LPC-file were developed to preserve the precision of query results. However, the performance of the vector approximation-based methods depend largely on the size of the approximation file and they lose the advantages of the multidimensional indexing methods that prune much search space. In this paper, we propose a new index structure called the GC-tree for efficient similarity search in image databases. The GC-tree is based on a special subspace partitioning strategy which is optimized for clustered high-dimensional images. It adaptively partitions the data space based on a density function and dynamically constructs an index structure. The resultant index structure adapts well to the strongly clustered distribution of high-dimensional images.

• Journal of Microbiology
• /
• v.40 no.4
• /
• pp.335-339
• /
• 2002

Partial sequences of 16S rRNA gene of five chroococcalian blue-green algal strains, Aphanothece nidulans KCTC AG10041, Aphanothece naegelii KCTC AG10042, Microcystis aeruginosa KCTC AG10159, Microcystis ichthyoblabe KCTC AG10160, and Microcystis viridis KCTC AG10198, which were isolated from water from the Geumgang River, were determined and were inferred their phylogenetic and taxonomic positions among taxa of order Chroococcales. Most taxa of Chroococcales whose partial 16S rRNA gene sequences were aligned in this study, are clustered with other related taxa. Aphanothece nidulans KCTC AG10041 and Aphanothece naegelii KCTC AG10042 made a cluster with other European species of these genera, which supported 100% of the bootstrap trees with a very high sequence similarity (97.4-99.4%) in this study. Three strains, Microcystis aeruginosa KCTC AG10159, M. ichthyoblabe KCTC AG10160, and M. viridis KCTC AG10198, formed a cluster with other Microcystis spp. supported 100 % of the bootstrap trees with a similarity of 97.0-99.9% except for two strains. However, this phylogentic tree made no resolution among the species of Microcystis spp. The topology of the tree reconfirmed the taxonomic status of three species of Microcystis, identified in this study based on the morphology, as three colonial types of Microcystis aeruginosa com. nov. Otsuka et al. (1999c). The genera of chroococcalian cyanophytes are heterogeneously clustered in these sequence analyses. We suggest that more molecular studies on the genera of Chroococcales with reference strains, widely collected from restricted geographic or environmental ranges, get accurate taxonomic or phylogenetic determinations.

• Proceedings of the Korean Society for Applied Microbiology Conference
• /
• 2004.06a
• /
• pp.258-265
• /
• 2004

Numerous genes of Escherichia coli have been shown to growth phase-dependent expression throughout growth. The global patterns of growth phase-dependent gene expression of E. coli throughout growth using oligonucleotide microarrays containing a nearly complete set of 4,289 annotated open reading frames. To determine the change of gene expression throughout growth, we compared RNAs taken from timecourses with common reference RNA, which is combined with equal amount of RNA pooled from each time point. The hierarchical clustering of the conditions in accordance with timecourse expression revealed that growth phases were clustered into four classes, consistent with known physiological growth status. We analyzed the differences of expression levels at genome level in both exponential and stationary growth phase cultures. Statistical analysis showed that 213 genes are shown to, growth phase-dependent expression. We also analyzed the expression of 256 known operons and 208 regulatory genes. To assess the global impact of RpoS, we identified 193 genes coregulated with rpoS and their expression levels were examined in the isogenic rpoS mutant. The results revealed that 99 of 193 were novel RpoS-dependent stationary phase-induced genes and the majority of those are functionally unknown. Our data provide that global changes and adjustments of gene expression are coordinately regulated by growth transition in E. coli.

• Fashion & Textile Research Journal
• /
• v.23 no.4
• /
• pp.480-490
• /
• 2021

Sales forecasting is crucial for many retail operations. For apparel retailers, accurate sales forecast for the next season is critical to properly manage inventory and plan their supply chains. The challenge in this increases because apparel products are always new for the next season, have numerous variations, short life cycles, long lead times, and seasonal trends. In this study, a sales forecasting model is proposed for apparel products using machine learning techniques. The sales data pertaining to outerwear items for four years were collected from a Korean sports brand and filtered with outliers. Subsequently, the data were standardized by removing the effects of exogenous variables. The sales patterns of outerwear items were clustered by applying K-means clustering, and outerwear attributes associated with the specific sales-pattern type were determined by using a decision tree classifier. Six types of sales pattern clusters were derived and classified using a hybrid model of clustering and decision tree algorithm, and finally, the relationship between outerwear attributes and sales patterns was revealed. Each sales pattern can be used to predict stock-keeping-unit-level sales based on item attributes.

• Asian-Australasian Journal of Animal Sciences
• /
• v.22 no.8
• /
• pp.1085-1090
• /
• 2009

There are currently five primary breeds of Chinese gamecock, the Henan, Luxi, Tulufan, Xishuangbanna andZhangzhou. Though there is historical evidence of cockfighting in China dating as far back as 2,800 years, the origin and genetic relationships of these breeds are not well understood. We used sequence variation from the mtDNA cytb gene and control region (1,697 bp) to examine the domestication history and genetic relationship of the Chinese gamecock. From 75 samples (14-16 per breed) we found 34 haplotypes, and 45 variable nucleotides. Phylogenetic reconstruction indicated multiple origins of the gamecock breeds. The breeds in the north and center of China, Tulufan, Luxi and Henan, clustered together in a haplogroup and may have the same ancestor. However the southern breeds, Zhangzhou and Xishuangbanna clustered into two isolated haplogroups, suggesting another two origins of Chinese gamecock. Meanwhile, extensive admixture was also found because samples from different breeds, more or less, were always grouped together in the same clades. Based on these results, we discuss the possibilities of multiple origins of gamecock breeds, from both ancestral gamecocks as well as other domestic chickens and red jungle fowl.