O slideshow foi denunciado.
Utilizamos seu perfil e dados de atividades no LinkedIn para personalizar e exibir anúncios mais relevantes. Altere suas preferências de anúncios quando desejar.

Local Outlier Factor

818 visualizações

Publicada em

my data mining lab , university of Bonn 2015

Publicada em: Educação
  • Seja o primeiro a comentar

  • Seja a primeira pessoa a gostar disto

Local Outlier Factor

  1. 1. Data Mining Lab, Local Outlier Factor Amr Koura / Page 1 Supervisor: Sebastian Bothe Local Outlier FactorLocal Outlier Factor
  2. 2. Data Mining Lab, Local Outlier Factor Amr Koura / Page 2 Supervisor: Sebastian Bothe LabGoalLabGoal Implement Local Outlier factory Batch Mode. Implement Local Outlier factory Incremental Mode. Comparetwo modes. Integratecodeinto open sourceproject “RealKD”: https://bitbucket.org/realKD/
  3. 3. Data Mining Lab, Local Outlier Factor Amr Koura / Page 3 Supervisor: Sebastian Bothe MotivationMotivation http://www.dbs.ifi.lmu.de/Publikationen/Papers/LOF.pdf
  4. 4. Data Mining Lab, Local Outlier Factor Amr Koura / Page 4 Supervisor: Sebastian Bothe Local Outlier FactorLocal Outlier Factor reach−distk ( A , B)=max(d (B , A),k−distance(B)) lrd (A)= 1 ∑ B∈KNN (A) reach−distk ( A , B)/k LOF (A)= 1 k ∑ B∈KNN (A) lrd (B) lrd ( A) https://en.wikipedia.org/wiki/Local_outlier_factor
  5. 5. Data Mining Lab, Local Outlier Factor Amr Koura / Page 5 Supervisor: Sebastian Bothe DemoDemo
  6. 6. Data Mining Lab, Local Outlier Factor Amr Koura / Page 6 Supervisor: Sebastian Bothe
  7. 7. Data Mining Lab, Local Outlier Factor Amr Koura / Page 7 Supervisor: Sebastian Bothe Incremental Outlier FactorIncremental Outlier Factor Motivation: - infinitestream makesmemory constraints. - computational constraint for processing each stream item. Goal: - Sameperformanceasiterated static LOF algorithm. - efficient algorithm: insertion/Deletion should effect only limited number of nearest neighbours http://www-ai.cs.uni-dortmund.de/LEHRE/FACHPROJEKT/SS12/paper/outlier/pokrajac2007.pdf
  8. 8. Data Mining Lab, Local Outlier Factor Amr Koura / Page 8 Supervisor: Sebastian Bothe Incremental LOF AdditionIncremental LOF Addition Berlin;52.520;13.380 Hamburg;53.550;10.000 Munchen;48.140;11.580 Bonn;50.730;7.100 Koln;50.950;6.970 Frankfurt;50.120;8.680 Dortmund;51.510;7.480 Stuttgart;48.790;9.190 Essen;51.470;7.000 Cairo;30.3;31.14 Hurghada;27.15;33.50
  9. 9. Data Mining Lab, Local Outlier Factor Amr Koura / Page 9 Supervisor: Sebastian Bothe Incremental LOF AdditionIncremental LOF Addition
  10. 10. Data Mining Lab, Local Outlier Factor Amr Koura / Page 10 Supervisor: Sebastian Bothe Incremental LOF AdditionIncremental LOF Addition 1 1.1909475617292364 2 1.1956830856346556 3 0.9645631106850818 4 0.8029601477829005 5 0.7577540135599361 6 0.7377495644370516 7 0.7509608512974867 8 0.99956101138198 9 0.6943310060958396 10 3.7497548217312158 11 3.928514077815152 Now , lets add new Point="Alexandria;31.13;29.58"
  11. 11. Data Mining Lab, Local Outlier Factor Amr Koura / Page 11 Supervisor: Sebastian Bothe Incremental LOF AdditionIncremental LOF Addition 0 1 2 6 5 8 7 4 3 9 10 1 0 6 5 8 7 4 3 2 9 10 2 7 5 0 1 3 4 6 8 9 10 3 4 8 6 5 7 1 2 0 9 10 4 3 8 6 5 7 1 2 0 9 10 5 7 3 4 6 8 1 2 0 9 10 6 8 4 3 5 7 1 0 2 9 10 7 5 2 3 4 6 8 1 0 9 10 8 6 4 3 5 7 1 0 2 9 10 9 10 2 0 7 5 1 3 4 6 8 10 9 2 0 7 5 1 3 4 6 8 0 1 2 6 5 8 7 4 3 11 9 10 1 0 6 5 8 7 4 3 2 11 9 10 2 7 5 0 1 3 4 6 8 11 9 10 3 4 8 6 5 7 1 2 0 11 9 10 4 3 8 6 5 7 1 2 0 11 9 10 5 7 3 4 6 8 1 2 0 11 9 10 6 8 4 3 5 7 1 0 2 11 9 10 7 5 2 3 4 6 8 1 0 11 9 10 8 6 4 3 5 7 1 0 2 11 9 10 9 11 10 2 0 7 5 1 3 4 6 8 10 9 11 2 0 7 5 1 3 4 6 8 11 9 10 2 0 7 5 1 3 4 6 8
  12. 12. Data Mining Lab, Local Outlier Factor Amr Koura / Page 12 Supervisor: Sebastian Bothe Incremental LOF AdditionIncremental LOF Addition Cities9,10 haschangein their K-distance. According to: TheLRD for citiesexistsin K-NN of cities(9,10) should updated LRD List={9,10,2} According to , all citesthat hasany of cities {9,10,2} in their new nearest neighbour should updatethier LOF value. LOF List={9,10,2,0,7} lrd (A)= 1 ∑ B∈KNN (A) reach−distk ( A , B)/k LOF ( A)= 1 k ∑ B∈KNN ( A) lrd (B) lrd (A)
  13. 13. Data Mining Lab, Local Outlier Factor Amr Koura / Page 13 Supervisor: Sebastian Bothe Comparison between staticandincremental LOFComparison between staticandincremental LOF Running static LOF output: 1.1909475617292364 1.1956830856346556 0.9645631106850818 0.8029601477829005 0.7577540135599361 0.7377495644370516 0.7509608512974867 0.99956101138198 0.6943310060958396 2.3423102537190847 2.342310253719085 2.342310253719085 Running incremental LOF and addition output: 1.1909475617292364 1.1956830856346556 0.9645631106850818 0.8029601477829005 0.7577540135599361 0.7377495644370516 0.7509608512974867 0.99956101138198 0.6943310060958396 2.3423102537190847 2.342310253719085 2.342310253719085
  14. 14. Data Mining Lab, Local Outlier Factor Amr Koura / Page 14 Supervisor: Sebastian Bothe ConclusionConclusion Implementation of Batch incremental modehasdone. Batch modecodeisintegrated into theproject repository while pull request hasmadeto integrateit. Incremental LOF hasequivalent detection performanceas static LOF. Incremental LOF requireslesscomputation timethan time. Incremental LOF complexity isO(N log N)
  15. 15. Data Mining Lab, Local Outlier Factor Amr Koura / Page 15 Supervisor: Sebastian Bothe Thank you

×