-
Notifications
You must be signed in to change notification settings - Fork 81
/
Copy pathoutlier_detection.html
1192 lines (962 loc) · 71.5 KB
/
outlier_detection.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<!DOCTYPE html>
<html lang="en" data-content_root="../" >
<head>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" /><meta name="viewport" content="width=device-width, initial-scale=1" />
<meta property="og:title" content="2.7. Novelty and Outlier Detection" />
<meta property="og:type" content="website" />
<meta property="og:url" content="https://scikit-learn/stable/modules/outlier_detection.html" />
<meta property="og:site_name" content="scikit-learn" />
<meta property="og:description" content="Many applications require being able to decide whether a new observation belongs to the same distribution as existing observations (it is an inlier), or should be considered as different (it is an ..." />
<meta property="og:image" content="https://scikit-learn/stable/_images/sphx_glr_plot_anomaly_comparison_001.png" />
<meta property="og:image:alt" content="scikit-learn" />
<meta name="description" content="Many applications require being able to decide whether a new observation belongs to the same distribution as existing observations (it is an inlier), or should be considered as different (it is an ..." />
<title>2.7. Novelty and Outlier Detection — scikit-learn 1.6.1 documentation</title>
<script data-cfasync="false">
document.documentElement.dataset.mode = localStorage.getItem("mode") || "";
document.documentElement.dataset.theme = localStorage.getItem("theme") || "";
</script>
<!--
this give us a css class that will be invisible only if js is disabled
-->
<noscript>
<style>
.pst-js-only { display: none !important; }
</style>
</noscript>
<!-- Loaded before other Sphinx assets -->
<link href="../_static/styles/theme.css?digest=8878045cc6db502f8baf" rel="stylesheet" />
<link href="../_static/styles/pydata-sphinx-theme.css?digest=8878045cc6db502f8baf" rel="stylesheet" />
<link rel="stylesheet" type="text/css" href="../_static/pygments.css?v=a746c00c" />
<link rel="stylesheet" type="text/css" href="../_static/copybutton.css?v=76b2166b" />
<link rel="stylesheet" type="text/css" href="../_static/plot_directive.css" />
<link rel="stylesheet" type="text/css" href="https://fonts.googleapis.com/css?family=Vibur" />
<link rel="stylesheet" type="text/css" href="../_static/jupyterlite_sphinx.css?v=e3ca86de" />
<link rel="stylesheet" type="text/css" href="../_static/sg_gallery.css?v=d2d258e8" />
<link rel="stylesheet" type="text/css" href="../_static/sg_gallery-binder.css?v=f4aeca0c" />
<link rel="stylesheet" type="text/css" href="../_static/sg_gallery-dataframe.css?v=2082cf3c" />
<link rel="stylesheet" type="text/css" href="../_static/sg_gallery-rendered-html.css?v=1277b6f3" />
<link rel="stylesheet" type="text/css" href="../_static/sphinx-design.min.css?v=95c83b7e" />
<link rel="stylesheet" type="text/css" href="../_static/styles/colors.css?v=cc94ab7d" />
<link rel="stylesheet" type="text/css" href="../_static/styles/custom.css?v=d67e4bb0" />
<!-- So that users can add custom icons -->
<script src="../_static/scripts/fontawesome.js?digest=8878045cc6db502f8baf"></script>
<!-- Pre-loaded scripts that we'll load fully later -->
<link rel="preload" as="script" href="../_static/scripts/bootstrap.js?digest=8878045cc6db502f8baf" />
<link rel="preload" as="script" href="../_static/scripts/pydata-sphinx-theme.js?digest=8878045cc6db502f8baf" />
<script src="../_static/documentation_options.js?v=d6a008b6"></script>
<script src="../_static/doctools.js?v=9a2dae69"></script>
<script src="../_static/sphinx_highlight.js?v=dc90522c"></script>
<script src="../_static/clipboard.min.js?v=a7894cd8"></script>
<script src="../_static/copybutton.js?v=97f0b27d"></script>
<script src="../_static/jupyterlite_sphinx.js?v=d6bdf5f8"></script>
<script src="../_static/design-tabs.js?v=f930bc37"></script>
<script data-domain="scikit-learn.org" defer="defer" src="https://views.scientific-python.org/js/script.js"></script>
<script async="async" src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-chtml.js"></script>
<script>DOCUMENTATION_OPTIONS.pagename = 'modules/outlier_detection';</script>
<script>
DOCUMENTATION_OPTIONS.theme_version = '0.16.1';
DOCUMENTATION_OPTIONS.theme_switcher_json_url = 'https://scikit-learn.org/dev/_static/versions.json';
DOCUMENTATION_OPTIONS.theme_switcher_version_match = '1.6.1';
DOCUMENTATION_OPTIONS.show_version_warning_banner =
true;
</script>
<script src="../_static/scripts/dropdown.js?v=e2048168"></script>
<script src="../_static/scripts/version-switcher.js?v=a6dd8357"></script>
<script src="../_static/scripts/sg_plotly_resize.js?v=eeb41cab"></script>
<link rel="canonical" href="https://scikit-learn.org/stable/modules/outlier_detection.html" />
<link rel="icon" href="../_static/favicon.ico"/>
<link rel="author" title="About these documents" href="../about.html" />
<link rel="search" title="Search" href="../search.html" />
<link rel="next" title="2.8. Density Estimation" href="density.html" />
<link rel="prev" title="2.6. Covariance estimation" href="covariance.html" />
<meta name="viewport" content="width=device-width, initial-scale=1"/>
<meta name="docsearch:language" content="en"/>
<meta name="docsearch:version" content="1.6" />
</head>
<body data-bs-spy="scroll" data-bs-target=".bd-toc-nav" data-offset="180" data-bs-root-margin="0px 0px -60%" data-default-mode="">
<div id="pst-skip-link" class="skip-link d-print-none"><a href="#main-content">Skip to main content</a></div>
<div id="pst-scroll-pixel-helper"></div>
<button type="button" class="btn rounded-pill" id="pst-back-to-top">
<i class="fa-solid fa-arrow-up"></i>Back to top</button>
<dialog id="pst-search-dialog">
<form class="bd-search d-flex align-items-center"
action="../search.html"
method="get">
<i class="fa-solid fa-magnifying-glass"></i>
<input type="search"
class="form-control"
name="q"
placeholder="Search the docs ..."
aria-label="Search the docs ..."
autocomplete="off"
autocorrect="off"
autocapitalize="off"
spellcheck="false"/>
<span class="search-button__kbd-shortcut"><kbd class="kbd-shortcut__modifier">Ctrl</kbd>+<kbd>K</kbd></span>
</form>
</dialog>
<div class="pst-async-banner-revealer d-none">
<aside id="bd-header-version-warning" class="d-none d-print-none" aria-label="Version warning"></aside>
</div>
<header class="bd-header navbar navbar-expand-lg bd-navbar d-print-none">
<div class="bd-header__inner bd-page-width">
<button class="pst-navbar-icon sidebar-toggle primary-toggle" aria-label="Site navigation">
<span class="fa-solid fa-bars"></span>
</button>
<div class=" navbar-header-items__start">
<div class="navbar-item">
<a class="navbar-brand logo" href="../index.html">
<img src="../_static/scikit-learn-logo-small.png" class="logo__image only-light" alt="scikit-learn homepage"/>
<img src="../_static/scikit-learn-logo-small.png" class="logo__image only-dark pst-js-only" alt="scikit-learn homepage"/>
</a></div>
</div>
<div class=" navbar-header-items">
<div class="me-auto navbar-header-items__center">
<div class="navbar-item">
<nav>
<ul class="bd-navbar-elements navbar-nav">
<li class="nav-item ">
<a class="nav-link nav-internal" href="../install.html">
Install
</a>
</li>
<li class="nav-item current active">
<a class="nav-link nav-internal" href="../user_guide.html">
User Guide
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../api/index.html">
API
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../auto_examples/index.html">
Examples
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-external" href="https://blog.scikit-learn.org/">
Community
</a>
</li>
<li class="nav-item dropdown">
<button class="btn dropdown-toggle nav-item" type="button"
data-bs-toggle="dropdown" aria-expanded="false"
aria-controls="pst-nav-more-links">
More
</button>
<ul id="pst-nav-more-links" class="dropdown-menu">
<li class=" ">
<a class="nav-link dropdown-item nav-internal" href="../getting_started.html">
Getting Started
</a>
</li>
<li class=" ">
<a class="nav-link dropdown-item nav-internal" href="../whats_new.html">
Release History
</a>
</li>
<li class=" ">
<a class="nav-link dropdown-item nav-internal" href="../glossary.html">
Glossary
</a>
</li>
<li class=" ">
<a class="nav-link dropdown-item nav-external" href="https://scikit-learn.org/dev/developers/index.html">
Development
</a>
</li>
<li class=" ">
<a class="nav-link dropdown-item nav-internal" href="../faq.html">
FAQ
</a>
</li>
<li class=" ">
<a class="nav-link dropdown-item nav-internal" href="../support.html">
Support
</a>
</li>
<li class=" ">
<a class="nav-link dropdown-item nav-internal" href="../related_projects.html">
Related Projects
</a>
</li>
<li class=" ">
<a class="nav-link dropdown-item nav-internal" href="../roadmap.html">
Roadmap
</a>
</li>
<li class=" ">
<a class="nav-link dropdown-item nav-internal" href="../governance.html">
Governance
</a>
</li>
<li class=" ">
<a class="nav-link dropdown-item nav-internal" href="../about.html">
About us
</a>
</li>
</ul>
</li>
</ul>
</nav></div>
</div>
<div class="navbar-header-items__end">
<div class="navbar-item navbar-persistent--container">
<button class="btn btn-sm pst-navbar-icon search-button search-button__button pst-js-only" title="Search" aria-label="Search" data-bs-placement="bottom" data-bs-toggle="tooltip">
<i class="fa-solid fa-magnifying-glass fa-lg"></i>
</button>
</div>
<div class="navbar-item">
<button class="btn btn-sm nav-link pst-navbar-icon theme-switch-button pst-js-only" aria-label="Color mode" data-bs-title="Color mode" data-bs-placement="bottom" data-bs-toggle="tooltip">
<i class="theme-switch fa-solid fa-sun fa-lg" data-mode="light" title="Light"></i>
<i class="theme-switch fa-solid fa-moon fa-lg" data-mode="dark" title="Dark"></i>
<i class="theme-switch fa-solid fa-circle-half-stroke fa-lg" data-mode="auto" title="System Settings"></i>
</button></div>
<div class="navbar-item"><ul class="navbar-icon-links"
aria-label="Icon Links">
<li class="nav-item">
<a href="https://github.com/scikit-learn/scikit-learn" title="GitHub" class="nav-link pst-navbar-icon" rel="noopener" target="_blank" data-bs-toggle="tooltip" data-bs-placement="bottom"><i class="fa-brands fa-square-github fa-lg" aria-hidden="true"></i>
<span class="sr-only">GitHub</span></a>
</li>
</ul></div>
<div class="navbar-item">
<div class="version-switcher__container dropdown pst-js-only">
<button id="pst-version-switcher-button-2"
type="button"
class="version-switcher__button btn btn-sm dropdown-toggle"
data-bs-toggle="dropdown"
aria-haspopup="listbox"
aria-controls="pst-version-switcher-list-2"
aria-label="Version switcher list"
>
Choose version <!-- this text may get changed later by javascript -->
<span class="caret"></span>
</button>
<div id="pst-version-switcher-list-2"
class="version-switcher__menu dropdown-menu list-group-flush py-0"
role="listbox" aria-labelledby="pst-version-switcher-button-2">
<!-- dropdown will be populated by javascript on page load -->
</div>
</div></div>
</div>
</div>
<div class="navbar-persistent--mobile">
<button class="btn btn-sm pst-navbar-icon search-button search-button__button pst-js-only" title="Search" aria-label="Search" data-bs-placement="bottom" data-bs-toggle="tooltip">
<i class="fa-solid fa-magnifying-glass fa-lg"></i>
</button>
</div>
<button class="pst-navbar-icon sidebar-toggle secondary-toggle" aria-label="On this page">
<span class="fa-solid fa-outdent"></span>
</button>
</div>
</header>
<div class="bd-container">
<div class="bd-container__inner bd-page-width">
<dialog id="pst-primary-sidebar-modal"></dialog>
<div id="pst-primary-sidebar" class="bd-sidebar-primary bd-sidebar">
<div class="sidebar-header-items sidebar-primary__section">
<div class="sidebar-header-items__center">
<div class="navbar-item">
<nav>
<ul class="bd-navbar-elements navbar-nav">
<li class="nav-item ">
<a class="nav-link nav-internal" href="../install.html">
Install
</a>
</li>
<li class="nav-item current active">
<a class="nav-link nav-internal" href="../user_guide.html">
User Guide
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../api/index.html">
API
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../auto_examples/index.html">
Examples
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-external" href="https://blog.scikit-learn.org/">
Community
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../getting_started.html">
Getting Started
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../whats_new.html">
Release History
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../glossary.html">
Glossary
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-external" href="https://scikit-learn.org/dev/developers/index.html">
Development
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../faq.html">
FAQ
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../support.html">
Support
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../related_projects.html">
Related Projects
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../roadmap.html">
Roadmap
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../governance.html">
Governance
</a>
</li>
<li class="nav-item ">
<a class="nav-link nav-internal" href="../about.html">
About us
</a>
</li>
</ul>
</nav></div>
</div>
<div class="sidebar-header-items__end">
<div class="navbar-item">
<button class="btn btn-sm nav-link pst-navbar-icon theme-switch-button pst-js-only" aria-label="Color mode" data-bs-title="Color mode" data-bs-placement="bottom" data-bs-toggle="tooltip">
<i class="theme-switch fa-solid fa-sun fa-lg" data-mode="light" title="Light"></i>
<i class="theme-switch fa-solid fa-moon fa-lg" data-mode="dark" title="Dark"></i>
<i class="theme-switch fa-solid fa-circle-half-stroke fa-lg" data-mode="auto" title="System Settings"></i>
</button></div>
<div class="navbar-item"><ul class="navbar-icon-links"
aria-label="Icon Links">
<li class="nav-item">
<a href="https://github.com/scikit-learn/scikit-learn" title="GitHub" class="nav-link pst-navbar-icon" rel="noopener" target="_blank" data-bs-toggle="tooltip" data-bs-placement="bottom"><i class="fa-brands fa-square-github fa-lg" aria-hidden="true"></i>
<span class="sr-only">GitHub</span></a>
</li>
</ul></div>
<div class="navbar-item">
<div class="version-switcher__container dropdown pst-js-only">
<button id="pst-version-switcher-button-3"
type="button"
class="version-switcher__button btn btn-sm dropdown-toggle"
data-bs-toggle="dropdown"
aria-haspopup="listbox"
aria-controls="pst-version-switcher-list-3"
aria-label="Version switcher list"
>
Choose version <!-- this text may get changed later by javascript -->
<span class="caret"></span>
</button>
<div id="pst-version-switcher-list-3"
class="version-switcher__menu dropdown-menu list-group-flush py-0"
role="listbox" aria-labelledby="pst-version-switcher-button-3">
<!-- dropdown will be populated by javascript on page load -->
</div>
</div></div>
</div>
</div>
<div class="sidebar-primary-items__start sidebar-primary__section">
<div class="sidebar-primary-item">
<nav class="bd-docs-nav bd-links"
aria-label="Section Navigation">
<p class="bd-links__title" role="heading" aria-level="1">Section Navigation</p>
<div class="bd-toc-item navbar-nav"><ul class="current nav bd-sidenav">
<li class="toctree-l1 has-children"><a class="reference internal" href="../supervised_learning.html">1. Supervised learning</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="linear_model.html">1.1. Linear Models</a></li>
<li class="toctree-l2"><a class="reference internal" href="lda_qda.html">1.2. Linear and Quadratic Discriminant Analysis</a></li>
<li class="toctree-l2"><a class="reference internal" href="kernel_ridge.html">1.3. Kernel ridge regression</a></li>
<li class="toctree-l2"><a class="reference internal" href="svm.html">1.4. Support Vector Machines</a></li>
<li class="toctree-l2"><a class="reference internal" href="sgd.html">1.5. Stochastic Gradient Descent</a></li>
<li class="toctree-l2"><a class="reference internal" href="neighbors.html">1.6. Nearest Neighbors</a></li>
<li class="toctree-l2"><a class="reference internal" href="gaussian_process.html">1.7. Gaussian Processes</a></li>
<li class="toctree-l2"><a class="reference internal" href="cross_decomposition.html">1.8. Cross decomposition</a></li>
<li class="toctree-l2"><a class="reference internal" href="naive_bayes.html">1.9. Naive Bayes</a></li>
<li class="toctree-l2"><a class="reference internal" href="tree.html">1.10. Decision Trees</a></li>
<li class="toctree-l2"><a class="reference internal" href="ensemble.html">1.11. Ensembles: Gradient boosting, random forests, bagging, voting, stacking</a></li>
<li class="toctree-l2"><a class="reference internal" href="multiclass.html">1.12. Multiclass and multioutput algorithms</a></li>
<li class="toctree-l2"><a class="reference internal" href="feature_selection.html">1.13. Feature selection</a></li>
<li class="toctree-l2"><a class="reference internal" href="semi_supervised.html">1.14. Semi-supervised learning</a></li>
<li class="toctree-l2"><a class="reference internal" href="isotonic.html">1.15. Isotonic regression</a></li>
<li class="toctree-l2"><a class="reference internal" href="calibration.html">1.16. Probability calibration</a></li>
<li class="toctree-l2"><a class="reference internal" href="neural_networks_supervised.html">1.17. Neural network models (supervised)</a></li>
</ul>
</details></li>
<li class="toctree-l1 current active has-children"><a class="reference internal" href="../unsupervised_learning.html">2. Unsupervised learning</a><details open="open"><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul class="current">
<li class="toctree-l2"><a class="reference internal" href="mixture.html">2.1. Gaussian mixture models</a></li>
<li class="toctree-l2"><a class="reference internal" href="manifold.html">2.2. Manifold learning</a></li>
<li class="toctree-l2"><a class="reference internal" href="clustering.html">2.3. Clustering</a></li>
<li class="toctree-l2"><a class="reference internal" href="biclustering.html">2.4. Biclustering</a></li>
<li class="toctree-l2"><a class="reference internal" href="decomposition.html">2.5. Decomposing signals in components (matrix factorization problems)</a></li>
<li class="toctree-l2"><a class="reference internal" href="covariance.html">2.6. Covariance estimation</a></li>
<li class="toctree-l2 current active"><a class="current reference internal" href="#">2.7. Novelty and Outlier Detection</a></li>
<li class="toctree-l2"><a class="reference internal" href="density.html">2.8. Density Estimation</a></li>
<li class="toctree-l2"><a class="reference internal" href="neural_networks_unsupervised.html">2.9. Neural network models (unsupervised)</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../model_selection.html">3. Model selection and evaluation</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="cross_validation.html">3.1. Cross-validation: evaluating estimator performance</a></li>
<li class="toctree-l2"><a class="reference internal" href="grid_search.html">3.2. Tuning the hyper-parameters of an estimator</a></li>
<li class="toctree-l2"><a class="reference internal" href="classification_threshold.html">3.3. Tuning the decision threshold for class prediction</a></li>
<li class="toctree-l2"><a class="reference internal" href="model_evaluation.html">3.4. Metrics and scoring: quantifying the quality of predictions</a></li>
<li class="toctree-l2"><a class="reference internal" href="learning_curve.html">3.5. Validation curves: plotting scores to evaluate models</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../inspection.html">4. Inspection</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="partial_dependence.html">4.1. Partial Dependence and Individual Conditional Expectation plots</a></li>
<li class="toctree-l2"><a class="reference internal" href="permutation_importance.html">4.2. Permutation feature importance</a></li>
</ul>
</details></li>
<li class="toctree-l1"><a class="reference internal" href="../visualizations.html">5. Visualizations</a></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../data_transforms.html">6. Dataset transformations</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="compose.html">6.1. Pipelines and composite estimators</a></li>
<li class="toctree-l2"><a class="reference internal" href="feature_extraction.html">6.2. Feature extraction</a></li>
<li class="toctree-l2"><a class="reference internal" href="preprocessing.html">6.3. Preprocessing data</a></li>
<li class="toctree-l2"><a class="reference internal" href="impute.html">6.4. Imputation of missing values</a></li>
<li class="toctree-l2"><a class="reference internal" href="unsupervised_reduction.html">6.5. Unsupervised dimensionality reduction</a></li>
<li class="toctree-l2"><a class="reference internal" href="random_projection.html">6.6. Random Projection</a></li>
<li class="toctree-l2"><a class="reference internal" href="kernel_approximation.html">6.7. Kernel Approximation</a></li>
<li class="toctree-l2"><a class="reference internal" href="metrics.html">6.8. Pairwise metrics, Affinities and Kernels</a></li>
<li class="toctree-l2"><a class="reference internal" href="preprocessing_targets.html">6.9. Transforming the prediction target (<code class="docutils literal notranslate"><span class="pre">y</span></code>)</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../datasets.html">7. Dataset loading utilities</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="../datasets/toy_dataset.html">7.1. Toy datasets</a></li>
<li class="toctree-l2"><a class="reference internal" href="../datasets/real_world.html">7.2. Real world datasets</a></li>
<li class="toctree-l2"><a class="reference internal" href="../datasets/sample_generators.html">7.3. Generated datasets</a></li>
<li class="toctree-l2"><a class="reference internal" href="../datasets/loading_other_datasets.html">7.4. Loading other datasets</a></li>
</ul>
</details></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../computing.html">8. Computing with scikit-learn</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="../computing/scaling_strategies.html">8.1. Strategies to scale computationally: bigger data</a></li>
<li class="toctree-l2"><a class="reference internal" href="../computing/computational_performance.html">8.2. Computational Performance</a></li>
<li class="toctree-l2"><a class="reference internal" href="../computing/parallelism.html">8.3. Parallelism, resource management, and configuration</a></li>
</ul>
</details></li>
<li class="toctree-l1"><a class="reference internal" href="../model_persistence.html">9. Model persistence</a></li>
<li class="toctree-l1"><a class="reference internal" href="../common_pitfalls.html">10. Common pitfalls and recommended practices</a></li>
<li class="toctree-l1 has-children"><a class="reference internal" href="../dispatching.html">11. Dispatching</a><details><summary><span class="toctree-toggle" role="presentation"><i class="fa-solid fa-chevron-down"></i></span></summary><ul>
<li class="toctree-l2"><a class="reference internal" href="array_api.html">11.1. Array API support (experimental)</a></li>
</ul>
</details></li>
<li class="toctree-l1"><a class="reference internal" href="../machine_learning_map.html">12. Choosing the right estimator</a></li>
<li class="toctree-l1"><a class="reference internal" href="../presentations.html">13. External Resources, Videos and Talks</a></li>
</ul>
</div>
</nav></div>
</div>
<div class="sidebar-primary-items__end sidebar-primary__section">
</div>
</div>
<main id="main-content" class="bd-main" role="main">
<div class="bd-content">
<div class="bd-article-container">
<div class="bd-header-article d-print-none">
<div class="header-article-items header-article__inner">
<div class="header-article-items__start">
<div class="header-article-item">
<nav aria-label="Breadcrumb" class="d-print-none">
<ul class="bd-breadcrumbs">
<li class="breadcrumb-item breadcrumb-home">
<a href="../index.html" class="nav-link" aria-label="Home">
<i class="fa-solid fa-home"></i>
</a>
</li>
<li class="breadcrumb-item"><a href="../user_guide.html" class="nav-link">User Guide</a></li>
<li class="breadcrumb-item"><a href="../unsupervised_learning.html" class="nav-link"><span class="section-number">2. </span>Unsupervised learning</a></li>
<li class="breadcrumb-item active" aria-current="page"><span class="ellipsis"><span class="section-number">2.7. </span>Novelty and Outlier Detection</span></li>
</ul>
</nav>
</div>
</div>
</div>
</div>
<div id="searchbox"></div>
<article class="bd-article">
<section id="novelty-and-outlier-detection">
<span id="outlier-detection"></span><h1><span class="section-number">2.7. </span>Novelty and Outlier Detection<a class="headerlink" href="#novelty-and-outlier-detection" title="Link to this heading">#</a></h1>
<p>Many applications require being able to decide whether a new observation
belongs to the same distribution as existing observations (it is an
<em>inlier</em>), or should be considered as different (it is an <em>outlier</em>).
Often, this ability is used to clean real data sets. Two important
distinctions must be made:</p>
<dl class="field-list simple">
<dt class="field-odd">outlier detection<span class="colon">:</span></dt>
<dd class="field-odd"><p>The training data contains outliers which are defined as observations that
are far from the others. Outlier detection estimators thus try to fit the
regions where the training data is the most concentrated, ignoring the
deviant observations.</p>
</dd>
<dt class="field-even">novelty detection<span class="colon">:</span></dt>
<dd class="field-even"><p>The training data is not polluted by outliers and we are interested in
detecting whether a <strong>new</strong> observation is an outlier. In this context an
outlier is also called a novelty.</p>
</dd>
</dl>
<p>Outlier detection and novelty detection are both used for anomaly
detection, where one is interested in detecting abnormal or unusual
observations. Outlier detection is then also known as unsupervised anomaly
detection and novelty detection as semi-supervised anomaly detection. In the
context of outlier detection, the outliers/anomalies cannot form a
dense cluster as available estimators assume that the outliers/anomalies are
located in low density regions. On the contrary, in the context of novelty
detection, novelties/anomalies can form a dense cluster as long as they are in
a low density region of the training data, considered as normal in this
context.</p>
<p>The scikit-learn project provides a set of machine learning tools that
can be used both for novelty or outlier detection. This strategy is
implemented with objects learning in an unsupervised way from the data:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">estimator</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X_train</span><span class="p">)</span>
</pre></div>
</div>
<p>new observations can then be sorted as inliers or outliers with a
<code class="docutils literal notranslate"><span class="pre">predict</span></code> method:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">estimator</span><span class="o">.</span><span class="n">predict</span><span class="p">(</span><span class="n">X_test</span><span class="p">)</span>
</pre></div>
</div>
<p>Inliers are labeled 1, while outliers are labeled -1. The predict method
makes use of a threshold on the raw scoring function computed by the
estimator. This scoring function is accessible through the <code class="docutils literal notranslate"><span class="pre">score_samples</span></code>
method, while the threshold can be controlled by the <code class="docutils literal notranslate"><span class="pre">contamination</span></code>
parameter.</p>
<p>The <code class="docutils literal notranslate"><span class="pre">decision_function</span></code> method is also defined from the scoring function,
in such a way that negative values are outliers and non-negative ones are
inliers:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">estimator</span><span class="o">.</span><span class="n">decision_function</span><span class="p">(</span><span class="n">X_test</span><span class="p">)</span>
</pre></div>
</div>
<p>Note that <a class="reference internal" href="generated/sklearn.neighbors.LocalOutlierFactor.html#sklearn.neighbors.LocalOutlierFactor" title="sklearn.neighbors.LocalOutlierFactor"><code class="xref py py-class docutils literal notranslate"><span class="pre">neighbors.LocalOutlierFactor</span></code></a> does not support
<code class="docutils literal notranslate"><span class="pre">predict</span></code>, <code class="docutils literal notranslate"><span class="pre">decision_function</span></code> and <code class="docutils literal notranslate"><span class="pre">score_samples</span></code> methods by default
but only a <code class="docutils literal notranslate"><span class="pre">fit_predict</span></code> method, as this estimator was originally meant to
be applied for outlier detection. The scores of abnormality of the training
samples are accessible through the <code class="docutils literal notranslate"><span class="pre">negative_outlier_factor_</span></code> attribute.</p>
<p>If you really want to use <a class="reference internal" href="generated/sklearn.neighbors.LocalOutlierFactor.html#sklearn.neighbors.LocalOutlierFactor" title="sklearn.neighbors.LocalOutlierFactor"><code class="xref py py-class docutils literal notranslate"><span class="pre">neighbors.LocalOutlierFactor</span></code></a> for novelty
detection, i.e. predict labels or compute the score of abnormality of new
unseen data, you can instantiate the estimator with the <code class="docutils literal notranslate"><span class="pre">novelty</span></code> parameter
set to <code class="docutils literal notranslate"><span class="pre">True</span></code> before fitting the estimator. In this case, <code class="docutils literal notranslate"><span class="pre">fit_predict</span></code> is
not available.</p>
<div class="admonition warning">
<p class="admonition-title">Warning</p>
<p><strong>Novelty detection with Local Outlier Factor</strong></p>
<p>When <code class="docutils literal notranslate"><span class="pre">novelty</span></code> is set to <code class="docutils literal notranslate"><span class="pre">True</span></code> be aware that you must only use
<code class="docutils literal notranslate"><span class="pre">predict</span></code>, <code class="docutils literal notranslate"><span class="pre">decision_function</span></code> and <code class="docutils literal notranslate"><span class="pre">score_samples</span></code> on new unseen data
and not on the training samples as this would lead to wrong results.
I.e., the result of <code class="docutils literal notranslate"><span class="pre">predict</span></code> will not be the same as <code class="docutils literal notranslate"><span class="pre">fit_predict</span></code>.
The scores of abnormality of the training samples are always accessible
through the <code class="docutils literal notranslate"><span class="pre">negative_outlier_factor_</span></code> attribute.</p>
</div>
<p>The behavior of <a class="reference internal" href="generated/sklearn.neighbors.LocalOutlierFactor.html#sklearn.neighbors.LocalOutlierFactor" title="sklearn.neighbors.LocalOutlierFactor"><code class="xref py py-class docutils literal notranslate"><span class="pre">neighbors.LocalOutlierFactor</span></code></a> is summarized in the
following table.</p>
<div class="pst-scrollable-table-container"><table class="table">
<thead>
<tr class="row-odd"><th class="head"><p>Method</p></th>
<th class="head"><p>Outlier detection</p></th>
<th class="head"><p>Novelty detection</p></th>
</tr>
</thead>
<tbody>
<tr class="row-even"><td><p><code class="docutils literal notranslate"><span class="pre">fit_predict</span></code></p></td>
<td><p>OK</p></td>
<td><p>Not available</p></td>
</tr>
<tr class="row-odd"><td><p><code class="docutils literal notranslate"><span class="pre">predict</span></code></p></td>
<td><p>Not available</p></td>
<td><p>Use only on new data</p></td>
</tr>
<tr class="row-even"><td><p><code class="docutils literal notranslate"><span class="pre">decision_function</span></code></p></td>
<td><p>Not available</p></td>
<td><p>Use only on new data</p></td>
</tr>
<tr class="row-odd"><td><p><code class="docutils literal notranslate"><span class="pre">score_samples</span></code></p></td>
<td><p>Use <code class="docutils literal notranslate"><span class="pre">negative_outlier_factor_</span></code></p></td>
<td><p>Use only on new data</p></td>
</tr>
<tr class="row-even"><td><p><code class="docutils literal notranslate"><span class="pre">negative_outlier_factor_</span></code></p></td>
<td><p>OK</p></td>
<td><p>OK</p></td>
</tr>
</tbody>
</table>
</div>
<section id="overview-of-outlier-detection-methods">
<h2><span class="section-number">2.7.1. </span>Overview of outlier detection methods<a class="headerlink" href="#overview-of-outlier-detection-methods" title="Link to this heading">#</a></h2>
<p>A comparison of the outlier detection algorithms in scikit-learn. Local
Outlier Factor (LOF) does not show a decision boundary in black as it
has no predict method to be applied on new data when it is used for outlier
detection.</p>
<figure class="align-center">
<a class="reference external image-reference" href="../auto_examples/miscellaneous/plot_anomaly_comparison.html"><img alt="../_images/sphx_glr_plot_anomaly_comparison_001.png" src="../_images/sphx_glr_plot_anomaly_comparison_001.png" style="width: 700.0px; height: 625.0px;" />
</a>
</figure>
<p><a class="reference internal" href="generated/sklearn.ensemble.IsolationForest.html#sklearn.ensemble.IsolationForest" title="sklearn.ensemble.IsolationForest"><code class="xref py py-class docutils literal notranslate"><span class="pre">ensemble.IsolationForest</span></code></a> and <a class="reference internal" href="generated/sklearn.neighbors.LocalOutlierFactor.html#sklearn.neighbors.LocalOutlierFactor" title="sklearn.neighbors.LocalOutlierFactor"><code class="xref py py-class docutils literal notranslate"><span class="pre">neighbors.LocalOutlierFactor</span></code></a>
perform reasonably well on the data sets considered here.
The <a class="reference internal" href="generated/sklearn.svm.OneClassSVM.html#sklearn.svm.OneClassSVM" title="sklearn.svm.OneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">svm.OneClassSVM</span></code></a> is known to be sensitive to outliers and thus
does not perform very well for outlier detection. That being said, outlier
detection in high-dimension, or without any assumptions on the distribution
of the inlying data is very challenging. <a class="reference internal" href="generated/sklearn.svm.OneClassSVM.html#sklearn.svm.OneClassSVM" title="sklearn.svm.OneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">svm.OneClassSVM</span></code></a> may still
be used with outlier detection but requires fine-tuning of its hyperparameter
<code class="docutils literal notranslate"><span class="pre">nu</span></code> to handle outliers and prevent overfitting.
<a class="reference internal" href="generated/sklearn.linear_model.SGDOneClassSVM.html#sklearn.linear_model.SGDOneClassSVM" title="sklearn.linear_model.SGDOneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">linear_model.SGDOneClassSVM</span></code></a> provides an implementation of a
linear One-Class SVM with a linear complexity in the number of samples. This
implementation is here used with a kernel approximation technique to obtain
results similar to <a class="reference internal" href="generated/sklearn.svm.OneClassSVM.html#sklearn.svm.OneClassSVM" title="sklearn.svm.OneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">svm.OneClassSVM</span></code></a> which uses a Gaussian kernel
by default. Finally, <a class="reference internal" href="generated/sklearn.covariance.EllipticEnvelope.html#sklearn.covariance.EllipticEnvelope" title="sklearn.covariance.EllipticEnvelope"><code class="xref py py-class docutils literal notranslate"><span class="pre">covariance.EllipticEnvelope</span></code></a> assumes the data is
Gaussian and learns an ellipse. For more details on the different estimators
refer to the example
<a class="reference internal" href="../auto_examples/miscellaneous/plot_anomaly_comparison.html#sphx-glr-auto-examples-miscellaneous-plot-anomaly-comparison-py"><span class="std std-ref">Comparing anomaly detection algorithms for outlier detection on toy datasets</span></a> and the
sections hereunder.</p>
<p class="rubric">Examples</p>
<ul class="simple">
<li><p>See <a class="reference internal" href="../auto_examples/miscellaneous/plot_anomaly_comparison.html#sphx-glr-auto-examples-miscellaneous-plot-anomaly-comparison-py"><span class="std std-ref">Comparing anomaly detection algorithms for outlier detection on toy datasets</span></a>
for a comparison of the <a class="reference internal" href="generated/sklearn.svm.OneClassSVM.html#sklearn.svm.OneClassSVM" title="sklearn.svm.OneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">svm.OneClassSVM</span></code></a>, the
<a class="reference internal" href="generated/sklearn.ensemble.IsolationForest.html#sklearn.ensemble.IsolationForest" title="sklearn.ensemble.IsolationForest"><code class="xref py py-class docutils literal notranslate"><span class="pre">ensemble.IsolationForest</span></code></a>, the
<a class="reference internal" href="generated/sklearn.neighbors.LocalOutlierFactor.html#sklearn.neighbors.LocalOutlierFactor" title="sklearn.neighbors.LocalOutlierFactor"><code class="xref py py-class docutils literal notranslate"><span class="pre">neighbors.LocalOutlierFactor</span></code></a> and
<a class="reference internal" href="generated/sklearn.covariance.EllipticEnvelope.html#sklearn.covariance.EllipticEnvelope" title="sklearn.covariance.EllipticEnvelope"><code class="xref py py-class docutils literal notranslate"><span class="pre">covariance.EllipticEnvelope</span></code></a>.</p></li>
<li><p>See <a class="reference internal" href="../auto_examples/miscellaneous/plot_outlier_detection_bench.html#sphx-glr-auto-examples-miscellaneous-plot-outlier-detection-bench-py"><span class="std std-ref">Evaluation of outlier detection estimators</span></a>
for an example showing how to evaluate outlier detection estimators,
the <a class="reference internal" href="generated/sklearn.neighbors.LocalOutlierFactor.html#sklearn.neighbors.LocalOutlierFactor" title="sklearn.neighbors.LocalOutlierFactor"><code class="xref py py-class docutils literal notranslate"><span class="pre">neighbors.LocalOutlierFactor</span></code></a> and the
<a class="reference internal" href="generated/sklearn.ensemble.IsolationForest.html#sklearn.ensemble.IsolationForest" title="sklearn.ensemble.IsolationForest"><code class="xref py py-class docutils literal notranslate"><span class="pre">ensemble.IsolationForest</span></code></a>, using ROC curves from
<a class="reference internal" href="generated/sklearn.metrics.RocCurveDisplay.html#sklearn.metrics.RocCurveDisplay" title="sklearn.metrics.RocCurveDisplay"><code class="xref py py-class docutils literal notranslate"><span class="pre">metrics.RocCurveDisplay</span></code></a>.</p></li>
</ul>
</section>
<section id="novelty-detection">
<h2><span class="section-number">2.7.2. </span>Novelty Detection<a class="headerlink" href="#novelty-detection" title="Link to this heading">#</a></h2>
<p>Consider a data set of <span class="math notranslate nohighlight">\(n\)</span> observations from the same
distribution described by <span class="math notranslate nohighlight">\(p\)</span> features. Consider now that we
add one more observation to that data set. Is the new observation so
different from the others that we can doubt it is regular? (i.e. does
it come from the same distribution?) Or on the contrary, is it so
similar to the other that we cannot distinguish it from the original
observations? This is the question addressed by the novelty detection
tools and methods.</p>
<p>In general, it is about to learn a rough, close frontier delimiting
the contour of the initial observations distribution, plotted in
embedding <span class="math notranslate nohighlight">\(p\)</span>-dimensional space. Then, if further observations
lay within the frontier-delimited subspace, they are considered as
coming from the same population than the initial
observations. Otherwise, if they lay outside the frontier, we can say
that they are abnormal with a given confidence in our assessment.</p>
<p>The One-Class SVM has been introduced by Schölkopf et al. for that purpose
and implemented in the <a class="reference internal" href="svm.html#svm"><span class="std std-ref">Support Vector Machines</span></a> module in the
<a class="reference internal" href="generated/sklearn.svm.OneClassSVM.html#sklearn.svm.OneClassSVM" title="sklearn.svm.OneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">svm.OneClassSVM</span></code></a> object. It requires the choice of a
kernel and a scalar parameter to define a frontier. The RBF kernel is
usually chosen although there exists no exact formula or algorithm to
set its bandwidth parameter. This is the default in the scikit-learn
implementation. The <code class="docutils literal notranslate"><span class="pre">nu</span></code> parameter, also known as the margin of
the One-Class SVM, corresponds to the probability of finding a new,
but regular, observation outside the frontier.</p>
<p class="rubric">References</p>
<ul class="simple">
<li><p><a class="reference external" href="https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/tr-99-87.pdf">Estimating the support of a high-dimensional distribution</a>
Schölkopf, Bernhard, et al. Neural computation 13.7 (2001): 1443-1471.</p></li>
</ul>
<p class="rubric">Examples</p>
<ul class="simple">
<li><p>See <a class="reference internal" href="../auto_examples/svm/plot_oneclass.html#sphx-glr-auto-examples-svm-plot-oneclass-py"><span class="std std-ref">One-class SVM with non-linear kernel (RBF)</span></a> for visualizing the
frontier learned around some data by a <a class="reference internal" href="generated/sklearn.svm.OneClassSVM.html#sklearn.svm.OneClassSVM" title="sklearn.svm.OneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">svm.OneClassSVM</span></code></a> object.</p></li>
<li><p><a class="reference internal" href="../auto_examples/applications/plot_species_distribution_modeling.html#sphx-glr-auto-examples-applications-plot-species-distribution-modeling-py"><span class="std std-ref">Species distribution modeling</span></a></p></li>
</ul>
<figure class="align-center">
<a class="reference external image-reference" href="../auto_examples/svm/plot_oneclass.html"><img alt="../_images/sphx_glr_plot_oneclass_001.png" src="../_images/sphx_glr_plot_oneclass_001.png" style="width: 480.0px; height: 360.0px;" />
</a>
</figure>
<section id="scaling-up-the-one-class-svm">
<h3><span class="section-number">2.7.2.1. </span>Scaling up the One-Class SVM<a class="headerlink" href="#scaling-up-the-one-class-svm" title="Link to this heading">#</a></h3>
<p>An online linear version of the One-Class SVM is implemented in
<a class="reference internal" href="generated/sklearn.linear_model.SGDOneClassSVM.html#sklearn.linear_model.SGDOneClassSVM" title="sklearn.linear_model.SGDOneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">linear_model.SGDOneClassSVM</span></code></a>. This implementation scales linearly with
the number of samples and can be used with a kernel approximation to
approximate the solution of a kernelized <a class="reference internal" href="generated/sklearn.svm.OneClassSVM.html#sklearn.svm.OneClassSVM" title="sklearn.svm.OneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">svm.OneClassSVM</span></code></a> whose
complexity is at best quadratic in the number of samples. See section
<a class="reference internal" href="sgd.html#sgd-online-one-class-svm"><span class="std std-ref">Online One-Class SVM</span></a> for more details.</p>
<p class="rubric">Examples</p>
<ul class="simple">
<li><p>See <a class="reference internal" href="../auto_examples/linear_model/plot_sgdocsvm_vs_ocsvm.html#sphx-glr-auto-examples-linear-model-plot-sgdocsvm-vs-ocsvm-py"><span class="std std-ref">One-Class SVM versus One-Class SVM using Stochastic Gradient Descent</span></a>
for an illustration of the approximation of a kernelized One-Class SVM
with the <code class="docutils literal notranslate"><span class="pre">linear_model.SGDOneClassSVM</span></code> combined with kernel approximation.</p></li>
</ul>
</section>
</section>
<section id="id1">
<h2><span class="section-number">2.7.3. </span>Outlier Detection<a class="headerlink" href="#id1" title="Link to this heading">#</a></h2>
<p>Outlier detection is similar to novelty detection in the sense that
the goal is to separate a core of regular observations from some
polluting ones, called <em>outliers</em>. Yet, in the case of outlier
detection, we don’t have a clean data set representing the population
of regular observations that can be used to train any tool.</p>
<section id="fitting-an-elliptic-envelope">
<h3><span class="section-number">2.7.3.1. </span>Fitting an elliptic envelope<a class="headerlink" href="#fitting-an-elliptic-envelope" title="Link to this heading">#</a></h3>
<p>One common way of performing outlier detection is to assume that the
regular data come from a known distribution (e.g. data are Gaussian
distributed). From this assumption, we generally try to define the
“shape” of the data, and can define outlying observations as
observations which stand far enough from the fit shape.</p>
<p>The scikit-learn provides an object
<a class="reference internal" href="generated/sklearn.covariance.EllipticEnvelope.html#sklearn.covariance.EllipticEnvelope" title="sklearn.covariance.EllipticEnvelope"><code class="xref py py-class docutils literal notranslate"><span class="pre">covariance.EllipticEnvelope</span></code></a> that fits a robust covariance
estimate to the data, and thus fits an ellipse to the central data
points, ignoring points outside the central mode.</p>
<p>For instance, assuming that the inlier data are Gaussian distributed, it
will estimate the inlier location and covariance in a robust way (i.e.
without being influenced by outliers). The Mahalanobis distances
obtained from this estimate is used to derive a measure of outlyingness.
This strategy is illustrated below.</p>
<figure class="align-center">
<a class="reference external image-reference" href="../auto_examples/covariance/plot_mahalanobis_distances.html"><img alt="../_images/sphx_glr_plot_mahalanobis_distances_001.png" src="../_images/sphx_glr_plot_mahalanobis_distances_001.png" style="width: 750.0px; height: 375.0px;" />
</a>
</figure>
<p class="rubric">Examples</p>
<ul class="simple">
<li><p>See <a class="reference internal" href="../auto_examples/covariance/plot_mahalanobis_distances.html#sphx-glr-auto-examples-covariance-plot-mahalanobis-distances-py"><span class="std std-ref">Robust covariance estimation and Mahalanobis distances relevance</span></a> for
an illustration of the difference between using a standard
(<a class="reference internal" href="generated/sklearn.covariance.EmpiricalCovariance.html#sklearn.covariance.EmpiricalCovariance" title="sklearn.covariance.EmpiricalCovariance"><code class="xref py py-class docutils literal notranslate"><span class="pre">covariance.EmpiricalCovariance</span></code></a>) or a robust estimate
(<a class="reference internal" href="generated/sklearn.covariance.MinCovDet.html#sklearn.covariance.MinCovDet" title="sklearn.covariance.MinCovDet"><code class="xref py py-class docutils literal notranslate"><span class="pre">covariance.MinCovDet</span></code></a>) of location and covariance to
assess the degree of outlyingness of an observation.</p></li>
<li><p>See <a class="reference internal" href="../auto_examples/applications/plot_outlier_detection_wine.html#sphx-glr-auto-examples-applications-plot-outlier-detection-wine-py"><span class="std std-ref">Outlier detection on a real data set</span></a>
for an example of robust covariance estimation on a real data set.</p></li>
</ul>
<p class="rubric">References</p>
<ul class="simple">
<li><p>Rousseeuw, P.J., Van Driessen, K. “A fast algorithm for the minimum
covariance determinant estimator” Technometrics 41(3), 212 (1999)</p></li>
</ul>
</section>
<section id="isolation-forest">
<span id="id2"></span><h3><span class="section-number">2.7.3.2. </span>Isolation Forest<a class="headerlink" href="#isolation-forest" title="Link to this heading">#</a></h3>
<p>One efficient way of performing outlier detection in high-dimensional datasets
is to use random forests.
The <a class="reference internal" href="generated/sklearn.ensemble.IsolationForest.html#sklearn.ensemble.IsolationForest" title="sklearn.ensemble.IsolationForest"><code class="xref py py-class docutils literal notranslate"><span class="pre">ensemble.IsolationForest</span></code></a> ‘isolates’ observations by randomly selecting
a feature and then randomly selecting a split value between the maximum and
minimum values of the selected feature.</p>
<p>Since recursive partitioning can be represented by a tree structure, the
number of splittings required to isolate a sample is equivalent to the path
length from the root node to the terminating node.</p>
<p>This path length, averaged over a forest of such random trees, is a
measure of normality and our decision function.</p>
<p>Random partitioning produces noticeably shorter paths for anomalies.
Hence, when a forest of random trees collectively produce shorter path
lengths for particular samples, they are highly likely to be anomalies.</p>
<p>The implementation of <a class="reference internal" href="generated/sklearn.ensemble.IsolationForest.html#sklearn.ensemble.IsolationForest" title="sklearn.ensemble.IsolationForest"><code class="xref py py-class docutils literal notranslate"><span class="pre">ensemble.IsolationForest</span></code></a> is based on an ensemble
of <a class="reference internal" href="generated/sklearn.tree.ExtraTreeRegressor.html#sklearn.tree.ExtraTreeRegressor" title="sklearn.tree.ExtraTreeRegressor"><code class="xref py py-class docutils literal notranslate"><span class="pre">tree.ExtraTreeRegressor</span></code></a>. Following Isolation Forest original paper,
the maximum depth of each tree is set to <span class="math notranslate nohighlight">\(\lceil \log_2(n) \rceil\)</span> where
<span class="math notranslate nohighlight">\(n\)</span> is the number of samples used to build the tree (see (Liu et al.,
2008) for more details).</p>
<p>This algorithm is illustrated below.</p>
<figure class="align-center">
<a class="reference external image-reference" href="../auto_examples/ensemble/plot_isolation_forest.html"><img alt="../_images/sphx_glr_plot_isolation_forest_003.png" src="../_images/sphx_glr_plot_isolation_forest_003.png" style="width: 480.0px; height: 360.0px;" />
</a>
</figure>
<p id="iforest-warm-start">The <a class="reference internal" href="generated/sklearn.ensemble.IsolationForest.html#sklearn.ensemble.IsolationForest" title="sklearn.ensemble.IsolationForest"><code class="xref py py-class docutils literal notranslate"><span class="pre">ensemble.IsolationForest</span></code></a> supports <code class="docutils literal notranslate"><span class="pre">warm_start=True</span></code> which
allows you to add more trees to an already fitted model:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="kn">from</span> <span class="nn">sklearn.ensemble</span> <span class="kn">import</span> <span class="n">IsolationForest</span>
<span class="gp">>>> </span><span class="kn">import</span> <span class="nn">numpy</span> <span class="k">as</span> <span class="nn">np</span>
<span class="gp">>>> </span><span class="n">X</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">array</span><span class="p">([[</span><span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">],</span> <span class="p">[</span><span class="o">-</span><span class="mi">2</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">],</span> <span class="p">[</span><span class="o">-</span><span class="mi">3</span><span class="p">,</span> <span class="o">-</span><span class="mi">2</span><span class="p">],</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">0</span><span class="p">],</span> <span class="p">[</span><span class="o">-</span><span class="mi">20</span><span class="p">,</span> <span class="mi">50</span><span class="p">],</span> <span class="p">[</span><span class="mi">3</span><span class="p">,</span> <span class="mi">5</span><span class="p">]])</span>
<span class="gp">>>> </span><span class="n">clf</span> <span class="o">=</span> <span class="n">IsolationForest</span><span class="p">(</span><span class="n">n_estimators</span><span class="o">=</span><span class="mi">10</span><span class="p">,</span> <span class="n">warm_start</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">clf</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X</span><span class="p">)</span> <span class="c1"># fit 10 trees </span>
<span class="gp">>>> </span><span class="n">clf</span><span class="o">.</span><span class="n">set_params</span><span class="p">(</span><span class="n">n_estimators</span><span class="o">=</span><span class="mi">20</span><span class="p">)</span> <span class="c1"># add 10 more trees </span>
<span class="gp">>>> </span><span class="n">clf</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">X</span><span class="p">)</span> <span class="c1"># fit the added trees </span>
</pre></div>
</div>
<p class="rubric">Examples</p>
<ul class="simple">
<li><p>See <a class="reference internal" href="../auto_examples/ensemble/plot_isolation_forest.html#sphx-glr-auto-examples-ensemble-plot-isolation-forest-py"><span class="std std-ref">IsolationForest example</span></a> for
an illustration of the use of IsolationForest.</p></li>
<li><p>See <a class="reference internal" href="../auto_examples/miscellaneous/plot_anomaly_comparison.html#sphx-glr-auto-examples-miscellaneous-plot-anomaly-comparison-py"><span class="std std-ref">Comparing anomaly detection algorithms for outlier detection on toy datasets</span></a>
for a comparison of <a class="reference internal" href="generated/sklearn.ensemble.IsolationForest.html#sklearn.ensemble.IsolationForest" title="sklearn.ensemble.IsolationForest"><code class="xref py py-class docutils literal notranslate"><span class="pre">ensemble.IsolationForest</span></code></a> with
<a class="reference internal" href="generated/sklearn.neighbors.LocalOutlierFactor.html#sklearn.neighbors.LocalOutlierFactor" title="sklearn.neighbors.LocalOutlierFactor"><code class="xref py py-class docutils literal notranslate"><span class="pre">neighbors.LocalOutlierFactor</span></code></a>,
<a class="reference internal" href="generated/sklearn.svm.OneClassSVM.html#sklearn.svm.OneClassSVM" title="sklearn.svm.OneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">svm.OneClassSVM</span></code></a> (tuned to perform like an outlier detection
method), <a class="reference internal" href="generated/sklearn.linear_model.SGDOneClassSVM.html#sklearn.linear_model.SGDOneClassSVM" title="sklearn.linear_model.SGDOneClassSVM"><code class="xref py py-class docutils literal notranslate"><span class="pre">linear_model.SGDOneClassSVM</span></code></a>, and a covariance-based
outlier detection with <a class="reference internal" href="generated/sklearn.covariance.EllipticEnvelope.html#sklearn.covariance.EllipticEnvelope" title="sklearn.covariance.EllipticEnvelope"><code class="xref py py-class docutils literal notranslate"><span class="pre">covariance.EllipticEnvelope</span></code></a>.</p></li>
</ul>
<p class="rubric">References</p>
<ul class="simple">
<li><p>Liu, Fei Tony, Ting, Kai Ming and Zhou, Zhi-Hua. “Isolation forest.”
Data Mining, 2008. ICDM’08. Eighth IEEE International Conference on.</p></li>
</ul>
</section>
<section id="local-outlier-factor">
<span id="id3"></span><h3><span class="section-number">2.7.3.3. </span>Local Outlier Factor<a class="headerlink" href="#local-outlier-factor" title="Link to this heading">#</a></h3>
<p>Another efficient way to perform outlier detection on moderately high dimensional
datasets is to use the Local Outlier Factor (LOF) algorithm.</p>
<p>The <a class="reference internal" href="generated/sklearn.neighbors.LocalOutlierFactor.html#sklearn.neighbors.LocalOutlierFactor" title="sklearn.neighbors.LocalOutlierFactor"><code class="xref py py-class docutils literal notranslate"><span class="pre">neighbors.LocalOutlierFactor</span></code></a> (LOF) algorithm computes a score
(called local outlier factor) reflecting the degree of abnormality of the
observations.
It measures the local density deviation of a given data point with respect to
its neighbors. The idea is to detect the samples that have a substantially
lower density than their neighbors.</p>