2008, Volume 10, Issue 9
Strategic Study of CAE >> 2008, Volume 10, Issue 9
Outliers detection algorithm based on nonlinear data transformation
Department of Computer Science and Technology, Nanjing University of Science and Technology, Nanjing 210094, China
Next Previous
Abstract
The data dimension reduction is the main method that can enhance the outliers mining efficiency based on higher-dimension data set. A novel outlier detection algorithm is proposed after analyzing the advantages and disadvantages of the classical outlier mining algorithm in the paper.In this paper, we can transform nonlinear large-scale data into linear data in the feature space,and introduce a nonlinear data transformation to reduce data dimension. On the basis of each resulting vector,it determins whether the data is outlier data or not one by one. This paper shows that the algorithm is not only used to detect linear separable outlier data,but also used to detect nonlinear inseparable outlier data. This indicate that the algorithm has its obvious superiority.
Keywords
dimension reduction ; kernel function ; principal component ; outliers
Figures
图1
图2
图3
References
[ 1 ] 夏火松主编.数据仓库与数据挖掘技术[ M] .北京: 科学出版 社, 2004
[ 2 ] Beyer K, Goldstein J, Ramakri Shnan R, et al.When is nearest neighbor meaningful [ A] .Been C, Buneman P ed.Proceedings of the 7th Intimation Conference on Data Theory Lecture Notes In Computer Science 1 540 [ C] .Jerusalem: Spnnger, 1999.217 - 235
[ 3 ] Li Yajun.Reforming the theory of invariant moments for pattern recognition [ J] .Pattern Recognition, 1992 , 25 ( 7 ) : 723 -730 link1
[ 4 ] Scholkopf B, Smola A, Muller K R.Nonlinear component analysisas a kernel eigenvalue problem [ J] .Neural Computation, 1998 , 10 : 1299 -1319
[ 5 ] Giudici P.Applied Data Ming: Statistical Methods for Business and Industry [ M] .Beijing: Electronics Industry Press, 2004
[ 6 ] Suykens J A K, Gestel T V, Vandewalle J, et al.A Support Vector Machine formulation to PCA Analysis and Its Kernel Ver- sion [ R] .ESAT -SCD -SISTA Technical Report 2002 -68 , Belgium: Katholieke Universiteit Leuven, 2002
[ 7 ] The third international knowledge discovery and data mining tools competition dataset KDD99 -Cup [ EB /OL ] .http: //kdd.ics. uci.edu /databases /kddcup99 /kddcup99.html, 1999
[ 8 ] Pal N R, Bezdek J C.On cluster validity for the fuzzy c -mesns model [ J] .IEEE Trans Fuzzy System 1995 , 3 ( 3 ) : 370 -379 link1
京公网安备 11010502051620号
