intelligent database systems lab n.y.u.s.t. i. m. fast exact k nearest neighbors search using an...
TRANSCRIPT
![Page 1: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/1.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.
Fast exact k nearest neighbors search using an orthogonal search tree
Presenter : Chun-Ping Wu Authors :Yi-Ching Liaw, Maw-Lin Leou, Chien-Min Wu
PR 2010
國立雲林科技大學National Yunlin University of Science and Technology
1
![Page 2: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/2.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Outline
Motivation Objective Methodology Experiments Conclusion Comments
2
![Page 3: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/3.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Motivation
The finding process of k nearest neighbors for a query point using FSA(full search algorithm) is very time consuming.
Many algorithms want to reduce the computational complexity of the kNN finding process. Pre-created tree structure
3
![Page 4: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/4.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Motivation
For a big PAT(Principal Axis Search), the computation time to evaluate boundary points and projection values will be large.
4
![Page 5: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/5.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Objective
To reduce the computation time on evaluation boundary points and projection values in the kNN searching process for a query point.
The proposed method requires no boundary points and only little computation time on evaluating projection values in the kNN finding process.
5
![Page 6: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/6.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Methodology
The OST(orthogonal search tree) algorithm OST construction process
K Nearest neighbors
search using the OST
6
![Page 7: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/7.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Methodology
The OST construction process
7
1,2,3,4,5,6,7,8,9
1,2,3,4,5,6,7,8,9
1,2,3 4,5,6 7,8,9
1,2,3,4,5,6,7,8,9
1,2,3 4,5,6 7,8,9
1 2 3
![Page 8: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/8.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Methodology
K nearest neighbors search
using the orthogonal search tree
8
1,2,3,4,5,6,7,8,9
1,2,3 4,5,6 7,8,9
1 2 3 4 5 6 7 8 9
![Page 9: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/9.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Experiments
Example 1 Uniform Markov source
9
![Page 10: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/10.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Experiments
10
![Page 11: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/11.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Experiments
Example 2 auto-correlated data
11
![Page 12: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/12.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Experiments
Example 3 Clustered Gaussian data
12
![Page 13: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/13.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Experiments
Example 4 Data sets are codebook
generated using 6 real images.
13
![Page 14: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/14.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Experiments
Example 5 Statlog data set.
14
34% 39%
![Page 15: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/15.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Conclusion
1515
Experimental results show that the proposed method always spends less computation time to find the kNN for a query point than the other methods.
The proposed method will find the same results as those of the FSA(full search algorithm).
![Page 16: Intelligent Database Systems Lab N.Y.U.S.T. I. M. Fast exact k nearest neighbors search using an orthogonal search tree Presenter : Chun-Ping Wu Authors](https://reader036.vdocuments.net/reader036/viewer/2022062716/56649e0a5503460f94af21f7/html5/thumbnails/16.jpg)
Intelligent Database Systems Lab
N.Y.U.S.T.
I. M.Comments
1616
Advantage To reduce the computation of the kNN finding process.
Drawback Lack of illustrations
Application Classification