• Title/Summary/Keyword: Tree-like

Search Result 500, Processing Time 0.028 seconds

Learning Algorithm for Multiple Distribution Data using Haar-like Feature and Decision Tree (다중 분포 학습 모델을 위한 Haar-like Feature와 Decision Tree를 이용한 학습 알고리즘)

  • Kwak, Ju-Hyun;Woen, Il-Young;Lee, Chang-Hoon
    • KIPS Transactions on Software and Data Engineering
    • /
    • v.2 no.1
    • /
    • pp.43-48
    • /
    • 2013
  • Adaboost is widely used for Haar-like feature boosting algorithm in Face Detection. It shows very effective performance on single distribution model. But when detecting front and side face images at same time, Adaboost shows it's limitation on multiple distribution data because it uses linear combination of basic classifier. This paper suggest the HDCT, modified decision tree algorithm for Haar-like features. We still tested the performance of HDCT compared with Adaboost on multiple distributed image recognition.

Optimization of Tree-like Core Overlay in Hybrid-structured Application-layer Multicast

  • Weng, Jianguang;Zou, Xuelan;Wang, Minhong
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.6 no.12
    • /
    • pp.3117-3132
    • /
    • 2012
  • The tree topology in multicast systems has high transmission efficiency, low latency, but poor resilience to node failures. In our work, some nodes are selected as backbone nodes to construct a tree-like core overlay. Backbone nodes are reliable enough and have strong upload capacity as well, which is helpful to overcome the shortcomings of tree topology. The core overlay is organized into a spanning tree while the whole overlay is of mesh-like topology. This paper focuses on improving the performance of the application-layer multicast overlay by optimizing the core overlay which is periodically adjusted with the proposed optimization algorithm. Our approach is to construct the overlay tree based on the out-degree weighted reliability where the reliability of a node is weighted by its upload bandwidth (out-degree). There is no illegal solution during the evolution which ensures the evolution efficiency. Simulation results show that the proposed approach greatly enhances the reliability of the tree-like core overlay systems and achieves shorter delay simultaneously. Its reliability performance is better than the reliability-first algorithm and its delay is very close to that of the degree-first algorithm. The complexity of the proposed algorithm is acceptable for application. Therefore the proposed approach is efficient for the topology optimization of a real multicast overlay.

Design and Implementation of the Tree-like Multiplier

  • Song, Gi-Yong;Lee, Jae-jin;Lee, Ho-Jun;Song, Ho-Jeong
    • Proceedings of the IEEK Conference
    • /
    • 2000.07a
    • /
    • pp.371-374
    • /
    • 2000
  • This paper proposes a 16-bit ${\times}$ 16-bit multiplier for 2 twos-complement binary numbers with tree-like structure and implements it on a FPGA. The space and time complexity analysis shows that the 16-bit Tree-like multiplier represents lower circuit complexity and computes more quickly than both Booth array multiplier md Modified array multiplier.

  • PDF

SQMR-tree: An Efficient Hybrid Index Structure for Large Spatial Data (SQMR-tree: 대용량 공간 데이타를 위한 효율적인 하이브리드 인덱스 구조)

  • Shin, In-Su;Kim, Joung-Joon;Kang, Hong-Koo;Han, Ki-Joon
    • Spatial Information Research
    • /
    • v.19 no.4
    • /
    • pp.45-54
    • /
    • 2011
  • In this paper, we propose a hybrid index structure, called the SQMR-tree(Spatial Quad MR-tree) that can process spatial data efficiently by combining advantages of the MR-tree and the SQR-tree. The MR-tree is an extended R-tree using a mapping tree to access directly to leaf nodes of the R-tree and the SQR-tree is a combination of the SQ-tree(Spatial Quad-tree) which is an extended Quad-tree to process spatial objects with non-zero area and the R-tree which actually stores spatial objects and are associated with each leaf node of the SQ-tree. The SQMR-tree consists of the SQR-tree as the base structure and the mapping trees associated with each R-tree of the SQR-tree. Therefore, because spatial objects are distributedly inserted into several R-trees and only R-trees intersected with the query area are accessed to process spatial queries like the SQR-tree, the query processing cost of the SQMR-tree can be reduced. Moreover, the search performance of the SQMR-tree is improved by using the mapping trees to access directly to leaf nodes of the R-tree without tree traversal like the MR-tree. Finally, we proved superiority of the SQMR-tree through experiments.

Design and Implementation of Trajectory Riving Tree for Combined Queries in Moving Object Databases (이동체 데이타베이스에서 복합 질의를 위한 궤적 분할 트리의 설계 및 구현)

  • 임덕성;전봉기;홍봉희;조대수
    • Journal of KIISE:Databases
    • /
    • v.31 no.2
    • /
    • pp.150-162
    • /
    • 2004
  • Moving objects have characteristics that they change continuously their positions over time. The movement of moving objects should be stored on trajectories for processing past queries. Moving objects databases need to provide spatio-temporal index for handling moving objects queries like combined queries. Combined queries consist of a range query selecting trajectories within a specific range and a trajectory query extracting to parts of the whole trajectory. Access methods showing good performance in range queries have a shortcoming that the cost of processing trajectory Queries is high. On the other hand, trajectory-based index schemes like the TB-tree are not suitable for range queries because of high overlaps between index nodes. This paper proposes new TR(Trajectory Riving)-tree which is revised for efficiently processing the combined queries. This index scheme has several features like the trajectory preservation, the increase of the capacity of leaf nodes, and the logical trajectory riving in order to reduce dead space and high overlap between bounding boxes of nodes. In our Performance study, the number of node access for combined queries in TR-tree is about 25% less than the STR-tree and the TB-tree.

Electrical Tree Simulation by Fractal Theory (Fractal 이론을 이용한 전기 트리 시뮬레이션)

  • Shin, T.S.;Shin, D.W.;Kang, S.H.;Lee, B.Y.;Lim, K.J.
    • Proceedings of the KIEE Conference
    • /
    • 1997.07d
    • /
    • pp.1481-1484
    • /
    • 1997
  • This paper describes a electrical tree simulation by fractal theory. Tree patterns produced by computer simulation with random numbers were studied from the point of view of fractal dimension. Tree patterns have a variety of shapes such as branch-like, bush-like, and quasi-bush-like trees. The patterns are determined by origins and probability ratio. The fractal dimensions have been measured a function of discharge number.

  • PDF

The Separation of Time and Space Tree for Moving or Static Objects in Limited Region (제한된 영역에서의 이동 및 고정 객체를 위한 시공간 분할 트리)

  • Yoon Jong-sun;Park Hyun-ju
    • Journal of Information Technology Applications and Management
    • /
    • v.12 no.1
    • /
    • pp.111-123
    • /
    • 2005
  • Many indexing methods were proposed so that process moving object efficiently. Among them, indexing methods like the 3D R-tree treat temporal and spatial domain as the same. Actually, however. both domain had better process separately because of difference in character and unit. Especially in this paper we deal with limited region such as indoor environment since spatial domain is limited but temporal domain is grown. In this paper we present a novel indexing structure, namely STS-tree(Separation of Time and Space tree). based on limited region. STS-tree is a hybrid tree structure which consists of R-tree and one-dimensional TB-tree. The R-tree component indexes static object and spatial information such as topography of the space. The TB-tree component indexes moving object and temporal information.

  • PDF

An Index Data Structure for String Search in External Memory (외부 메모리에서 문자열을 효율적으로 탐색하기 위한 인덱스 자료 구조)

  • Na, Joong-Chae;Park, Kun-Soo
    • Journal of KIISE:Computer Systems and Theory
    • /
    • v.32 no.11_12
    • /
    • pp.598-607
    • /
    • 2005
  • We propose a new external-memory index data structure, the Suffix B-tree. The Suffix B-tree is a B-tree in which the key is a string like the String B-tree. While the node in the String B-tree is implemented with a Patricia trio, the node in the Suffix B-tree is implemented with an array. So the Suffix B-tree is simpler and easier to be Implemented than the String B-tree. Nevertheless, the branching algorithm of the Suffix B-tree is as efficient as that of the String B-tree. Consequently, the Suffix B-tree takes the same worst-case disk accesses as the String B-tree to solve the string matching problem, which is fundamental and important in the area of string algorithms.

Vertical Distribution of Foraging Tits in Mixed Species Flocks in Urban Forests

  • Lee, Sang-Don
    • The Korean Journal of Ecology
    • /
    • v.22 no.2
    • /
    • pp.65-68
    • /
    • 1999
  • In December-January of 1996-1997 and 1997-1998, information was gathered about vertical distribution of foraging sites of tits in 34 flocks in coniferous and deciduous forests. There was a significant effect of forest type on the distribution of foraging sites of each species. Habitat was classified into 5 height layers vertically: ground, bushes (usually<1.5 m, up to 3 m), tree layer 1 (up to 1/3 of tree height), tree layer 2 (1/3-2/3 tree height). and tree layer 3 (>2/3 tree height). There were differences among species: great tit (Parus major) foraged mostly on the ground, coal tit (P. ater) and long-tailed tit (Acrocephalus caudatus) - on the highest tree layer, marsh tit (P. palustris) was often seen on bushes, and varied tit (P. varius) - in tree layer 2. Smaller species used upper and outer parts of trees. suggesting that, like in most other similar studies. larger dominant species prevented smaller species from using inner parts of trees.

  • PDF

바이오 센서 응용을 위한 Tree-like 실리콘 나노와이어의 표면성장 및 특성파악

  • An, Chi-Seong;Kulkarni, Atul;Kim, Ho-Jung;Kim, Tae-Seong
    • Proceedings of the Korean Vacuum Society Conference
    • /
    • 2011.08a
    • /
    • pp.346-346
    • /
    • 2011
  • 실리콘 나노와이어는 높은 표면적으로 인해 뛰어난 감지 능력을 가지는 재료 중 하나로 다양한 센서 응용 분야에 사용되고 있다. 이를 제작하는 방법에는 Micro Electro Mechanical Systems (MEMS) 공정을 이용한 Top-down 방식과 Vapor-Liquid-Solid (VLS) 공정을 이용한 Bottom-up 방식이 널리 사용되고 있다. 특히 Plasma-Enhanced Chemical Vapor Deposition(PECVD)와 Au 촉매를 이용한 Bottom-up 방식은 수십 나노미터 이하의 실리콘 나노와이어를 간단한 변수 조절을 통해 성장시킬 수 있다. 또한 Au/Si의 공융점인 363$^{\circ}C$보다 낮은 온도에서 $SiH_4$를 분해시킬 수 있어 열적 효과로 인한 손실을 줄일 수 있는 장점을 지니고 있다. 하지만 PECVD를 이용한 실리콘 나노와이어 성장은 VLS 공정을 통해 표면으로부터 수직으로 성장하게 되는데 이는 센서 응용을 위한 전극 사이의 수평 연결 어려움을 지니고 있다. 따라서 이를 피하기 위한 표면 성장된 실리콘 나노와이어가 요구된다. 본 연구에서는 PECVD VLS 공정을 이용하여 $HAuCl_4$를 촉매로 이용한 표면 성장된 Tree-like 실리콘 나노와이어를 성장시켰다. 공정가스로는 $SiH_4$와 이를 분해시키기 위해 Ar 플라즈마를 사용 하였고 웨이퍼 표면에 HAuCl4를 분사하고 고진공 상태에서 챔버 기판을 370$^{\circ}C$까지 가열한 후 플라즈마 파워(W) 및 공정 압력(mTorr)을 변수로 두어 실험을 진행하였다. 기존의 보고된 연구와 달리 환원된 금 입자 대신 $HAuCl_4$용액을 그대로 사용하였는데 이는 표면 조도(Surface roughness)를 가지는 Au 박막 상태로 존재하게 된다. 이 중 마루(Asperite) 부분에 PECVD로부터 발생된 실리콘 나노 입자가 상대적으로 높은 확률로 흡착하게 되어 실리콘 나노와이어의 표면성장을 유도하게 된다. 성장된 실리콘 나노와이어는 SEM과 EDS를 이용하여 직경, 길이 및 화학적 성분을 측정하였다. 직경은 약 100 nm, 길이는 약 10 ${\mu}m$ 정도로 나타났으며 Tree-like 실리콘 나노와이어가 성장되었다. 향후 전극이 형성된 기판위에 이를 직접 성장시킴으로써 이 물질의 I-V 특성을 파악 할 것이며 이는 센서 응용 분야에 도움이 될 것으로 기대된다.

  • PDF