-
Notifications
You must be signed in to change notification settings - Fork 81
/
Copy pathmodel_evaluation.html
3991 lines (3751 loc) · 418 KB
/
model_evaluation.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="3.4. Metrics and scoring: quantifying the quality of predictions" />
<meta property="og:type" content="website" />
<meta property="og:url" content="https://scikit-learn/stable/modules/model_evaluation.html" />
<meta property="og:site_name" content="scikit-learn" />
<meta property="og:description" content="Which scoring function should I use?: Before we take a closer look into the details of the many scores and evaluation metrics, we want to give some guidance, inspired by statistical decision theory..." />
<meta property="og:image" content="https://scikit-learn/stable/_images/sphx_glr_plot_confusion_matrix_001.png" />
<meta property="og:image:alt" content="scikit-learn" />
<meta name="description" content="Which scoring function should I use?: Before we take a closer look into the details of the many scores and evaluation metrics, we want to give some guidance, inspired by statistical decision theory..." />
<title>3.4. Metrics and scoring: quantifying the quality of predictions — scikit-learn 1.7.dev0 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=8f2a1f02" />
<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=2c9f8f05" />
<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=8f525996" />
<!-- 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=473747f4"></script>
<script src="../_static/doctools.js?v=9bcbadda"></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=96e329c5"></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/model_evaluation';</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 = '1.7.dev0';
DOCUMENTATION_OPTIONS.show_version_warning_banner =
true;
</script>
<script src="../_static/scripts/dropdown.js?v=d6825577"></script>
<script src="../_static/scripts/version-switcher.js?v=a6dd8357"></script>
<script src="../_static/scripts/sg_plotly_resize.js?v=2167d4db"></script>
<link rel="canonical" href="https://scikit-learn.org/stable/modules/model_evaluation.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="3.5. Validation curves: plotting scores to evaluate models" href="learning_curve.html" />
<link rel="prev" title="3.3. Tuning the decision threshold for class prediction" href="classification_threshold.html" />
<meta name="viewport" content="width=device-width, initial-scale=1"/>
<meta name="docsearch:language" content="en"/>
<meta name="docsearch:version" content="1.7" />
</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-internal" href="../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-internal" href="../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 has-children"><a class="reference internal" href="../unsupervised_learning.html">2. Unsupervised 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="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"><a class="reference internal" href="outlier_detection.html">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 current active has-children"><a class="reference internal" href="../model_selection.html">3. Model selection and evaluation</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="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 current active"><a class="current reference internal" href="#">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="../model_selection.html" class="nav-link"><span class="section-number">3. </span>Model selection and evaluation</a></li>
<li class="breadcrumb-item active" aria-current="page"><span class="ellipsis"><span class="section-number">3.4. </span>Metrics and scoring: quantifying the quality of predictions</span></li>
</ul>
</nav>
</div>
</div>
</div>
</div>
<div id="searchbox"></div>
<article class="bd-article">
<section id="metrics-and-scoring-quantifying-the-quality-of-predictions">
<span id="model-evaluation"></span><h1><span class="section-number">3.4. </span>Metrics and scoring: quantifying the quality of predictions<a class="headerlink" href="#metrics-and-scoring-quantifying-the-quality-of-predictions" title="Link to this heading">#</a></h1>
<section id="which-scoring-function-should-i-use">
<span id="which-scoring-function"></span><h2><span class="section-number">3.4.1. </span>Which scoring function should I use?<a class="headerlink" href="#which-scoring-function-should-i-use" title="Link to this heading">#</a></h2>
<p>Before we take a closer look into the details of the many scores and
<a class="reference internal" href="../glossary.html#term-evaluation-metrics"><span class="xref std std-term">evaluation metrics</span></a>, we want to give some guidance, inspired by statistical
decision theory, on the choice of <strong>scoring functions</strong> for <strong>supervised learning</strong>,
see <a class="reference internal" href="#gneiting2009" id="id1"><span>[Gneiting2009]</span></a>:</p>
<ul class="simple">
<li><p><em>Which scoring function should I use?</em></p></li>
<li><p><em>Which scoring function is a good one for my task?</em></p></li>
</ul>
<p>In a nutshell, if the scoring function is given, e.g. in a kaggle competition
or in a business context, use that one.
If you are free to choose, it starts by considering the ultimate goal and application
of the prediction. It is useful to distinguish two steps:</p>
<ul class="simple">
<li><p>Predicting</p></li>
<li><p>Decision making</p></li>
</ul>
<p><strong>Predicting:</strong>
Usually, the response variable <span class="math notranslate nohighlight">\(Y\)</span> is a random variable, in the sense that there
is <em>no deterministic</em> function <span class="math notranslate nohighlight">\(Y = g(X)\)</span> of the features <span class="math notranslate nohighlight">\(X\)</span>.
Instead, there is a probability distribution <span class="math notranslate nohighlight">\(F\)</span> of <span class="math notranslate nohighlight">\(Y\)</span>.
One can aim to predict the whole distribution, known as <em>probabilistic prediction</em>,
or—more the focus of scikit-learn—issue a <em>point prediction</em> (or point forecast)
by choosing a property or functional of that distribution <span class="math notranslate nohighlight">\(F\)</span>.
Typical examples are the mean (expected value), the median or a quantile of the
response variable <span class="math notranslate nohighlight">\(Y\)</span> (conditionally on <span class="math notranslate nohighlight">\(X\)</span>).</p>
<p>Once that is settled, use a <strong>strictly consistent</strong> scoring function for that
(target) functional, see <a class="reference internal" href="#gneiting2009" id="id2"><span>[Gneiting2009]</span></a>.
This means using a scoring function that is aligned with <em>measuring the distance
between predictions</em> <code class="docutils literal notranslate"><span class="pre">y_pred</span></code> <em>and the true target functional using observations of</em>
<span class="math notranslate nohighlight">\(Y\)</span>, i.e. <code class="docutils literal notranslate"><span class="pre">y_true</span></code>.
For classification <strong>strictly proper scoring rules</strong>, see
<a class="reference external" href="https://en.wikipedia.org/wiki/Scoring_rule">Wikipedia entry for Scoring rule</a>
and <a class="reference internal" href="#gneiting2007" id="id3"><span>[Gneiting2007]</span></a>, coincide with strictly consistent scoring functions.
The table further below provides examples.
One could say that consistent scoring functions act as <em>truth serum</em> in that
they guarantee <em>“that truth telling […] is an optimal strategy in
expectation”</em> <a class="reference internal" href="#gneiting2014" id="id4"><span>[Gneiting2014]</span></a>.</p>
<p>Once a strictly consistent scoring function is chosen, it is best used for both: as
loss function for model training and as metric/score in model evaluation and model
comparison.</p>
<p>Note that for regressors, the prediction is done with <a class="reference internal" href="../glossary.html#term-predict"><span class="xref std std-term">predict</span></a> while for
classifiers it is usually <a class="reference internal" href="../glossary.html#term-predict_proba"><span class="xref std std-term">predict_proba</span></a>.</p>
<p><strong>Decision Making:</strong>
The most common decisions are done on binary classification tasks, where the result of
<a class="reference internal" href="../glossary.html#term-predict_proba"><span class="xref std std-term">predict_proba</span></a> is turned into a single outcome, e.g., from the predicted
probability of rain a decision is made on how to act (whether to take mitigating
measures like an umbrella or not).
For classifiers, this is what <a class="reference internal" href="../glossary.html#term-predict"><span class="xref std std-term">predict</span></a> returns.
See also <a class="reference internal" href="classification_threshold.html#tunedthresholdclassifiercv"><span class="std std-ref">Tuning the decision threshold for class prediction</span></a>.
There are many scoring functions which measure different aspects of such a
decision, most of them are covered with or derived from the
<a class="reference internal" href="generated/sklearn.metrics.confusion_matrix.html#sklearn.metrics.confusion_matrix" title="sklearn.metrics.confusion_matrix"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.confusion_matrix</span></code></a>.</p>
<p><strong>List of strictly consistent scoring functions:</strong>
Here, we list some of the most relevant statistical functionals and corresponding
strictly consistent scoring functions for tasks in practice. Note that the list is not
complete and that there are more of them.
For further criteria on how to select a specific one, see <a class="reference internal" href="#fissler2022" id="id5"><span>[Fissler2022]</span></a>.</p>
<div class="pst-scrollable-table-container"><table class="table">
<thead>
<tr class="row-odd"><th class="head"><p>functional</p></th>
<th class="head"><p>scoring or loss function</p></th>
<th class="head"><p>response <code class="docutils literal notranslate"><span class="pre">y</span></code></p></th>
<th class="head"><p>prediction</p></th>
</tr>
</thead>
<tbody>
<tr class="row-even"><td><p><strong>Classification</strong></p></td>
<td></td>
<td></td>
<td></td>
</tr>
<tr class="row-odd"><td><p>mean</p></td>
<td><p><a class="reference internal" href="#brier-score-loss"><span class="std std-ref">Brier score</span></a> <sup>1</sup></p></td>
<td><p>multi-class</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">predict_proba</span></code></p></td>
</tr>
<tr class="row-even"><td><p>mean</p></td>
<td><p><a class="reference internal" href="#log-loss"><span class="std std-ref">log loss</span></a></p></td>
<td><p>multi-class</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">predict_proba</span></code></p></td>
</tr>
<tr class="row-odd"><td><p>mode</p></td>
<td><p><a class="reference internal" href="#zero-one-loss"><span class="std std-ref">zero-one loss</span></a> <sup>2</sup></p></td>
<td><p>multi-class</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">predict</span></code>, categorical</p></td>
</tr>
<tr class="row-even"><td><p><strong>Regression</strong></p></td>
<td></td>
<td></td>
<td></td>
</tr>
<tr class="row-odd"><td><p>mean</p></td>
<td><p><a class="reference internal" href="#mean-squared-error"><span class="std std-ref">squared error</span></a> <sup>3</sup></p></td>
<td><p>all reals</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">predict</span></code>, all reals</p></td>
</tr>
<tr class="row-even"><td><p>mean</p></td>
<td><p><a class="reference internal" href="#mean-tweedie-deviance"><span class="std std-ref">Poisson deviance</span></a></p></td>
<td><p>non-negative</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">predict</span></code>, strictly positive</p></td>
</tr>
<tr class="row-odd"><td><p>mean</p></td>
<td><p><a class="reference internal" href="#mean-tweedie-deviance"><span class="std std-ref">Gamma deviance</span></a></p></td>
<td><p>strictly positive</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">predict</span></code>, strictly positive</p></td>
</tr>
<tr class="row-even"><td><p>mean</p></td>
<td><p><a class="reference internal" href="#mean-tweedie-deviance"><span class="std std-ref">Tweedie deviance</span></a></p></td>
<td><p>depends on <code class="docutils literal notranslate"><span class="pre">power</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">predict</span></code>, depends on <code class="docutils literal notranslate"><span class="pre">power</span></code></p></td>
</tr>
<tr class="row-odd"><td><p>median</p></td>
<td><p><a class="reference internal" href="#mean-absolute-error"><span class="std std-ref">absolute error</span></a></p></td>
<td><p>all reals</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">predict</span></code>, all reals</p></td>
</tr>
<tr class="row-even"><td><p>quantile</p></td>
<td><p><a class="reference internal" href="#pinball-loss"><span class="std std-ref">pinball loss</span></a></p></td>
<td><p>all reals</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">predict</span></code>, all reals</p></td>
</tr>
<tr class="row-odd"><td><p>mode</p></td>
<td><p>no consistent one exists</p></td>
<td><p>reals</p></td>
<td></td>
</tr>
</tbody>
</table>
</div>
<p><sup>1</sup> The Brier score is just a different name for the squared error in case of
classification.</p>
<p><sup>2</sup> The zero-one loss is only consistent but not strictly consistent for the mode.
The zero-one loss is equivalent to one minus the accuracy score, meaning it gives
different score values but the same ranking.</p>
<p><sup>3</sup> R² gives the same ranking as squared error.</p>
<p><strong>Fictitious Example:</strong>
Let’s make the above arguments more tangible. Consider a setting in network reliability
engineering, such as maintaining stable internet or Wi-Fi connections.
As provider of the network, you have access to the dataset of log entries of network
connections containing network load over time and many interesting features.
Your goal is to improve the reliability of the connections.
In fact, you promise your customers that on at least 99% of all days there are no
connection discontinuities larger than 1 minute.
Therefore, you are interested in a prediction of the 99% quantile (of longest
connection interruption duration per day) in order to know in advance when to add
more bandwidth and thereby satisfy your customers. So the <em>target functional</em> is the
99% quantile. From the table above, you choose the pinball loss as scoring function
(fair enough, not much choice given), for model training (e.g.
<code class="docutils literal notranslate"><span class="pre">HistGradientBoostingRegressor(loss="quantile",</span> <span class="pre">quantile=0.99)</span></code>) as well as model
evaluation (<code class="docutils literal notranslate"><span class="pre">mean_pinball_loss(...,</span> <span class="pre">alpha=0.99)</span></code> - we apologize for the different
argument names, <code class="docutils literal notranslate"><span class="pre">quantile</span></code> and <code class="docutils literal notranslate"><span class="pre">alpha</span></code>) be it in grid search for finding
hyperparameters or in comparing to other models like
<code class="docutils literal notranslate"><span class="pre">QuantileRegressor(quantile=0.99)</span></code>.</p>
<p class="rubric">References</p>
<div role="list" class="citation-list">
<div class="citation" id="gneiting2007" role="doc-biblioentry">
<span class="label"><span class="fn-bracket">[</span><a role="doc-backlink" href="#id3">Gneiting2007</a><span class="fn-bracket">]</span></span>
<p>T. Gneiting and A. E. Raftery. <a class="reference external" href="https://doi.org/10.1198/016214506000001437">Strictly Proper
Scoring Rules, Prediction, and Estimation</a>
In: Journal of the American Statistical Association 102 (2007),
pp. 359– 378.
<a class="reference external" href="https://sites.stat.washington.edu/raftery/Research/PDF/Gneiting2007jasa.pdf">link to pdf</a></p>
</div>
<div class="citation" id="gneiting2009" role="doc-biblioentry">
<span class="label"><span class="fn-bracket">[</span>Gneiting2009<span class="fn-bracket">]</span></span>
<span class="backrefs">(<a role="doc-backlink" href="#id1">1</a>,<a role="doc-backlink" href="#id2">2</a>)</span>
<p>T. Gneiting. <a class="reference external" href="https://arxiv.org/abs/0912.0902">Making and Evaluating Point Forecasts</a>
Journal of the American Statistical Association 106 (2009): 746 - 762.</p>
</div>
<div class="citation" id="gneiting2014" role="doc-biblioentry">
<span class="label"><span class="fn-bracket">[</span><a role="doc-backlink" href="#id4">Gneiting2014</a><span class="fn-bracket">]</span></span>
<p>T. Gneiting and M. Katzfuss. <a class="reference external" href="https://doi.org/10.1146/annurev-statistics-062713-085831">Probabilistic Forecasting</a>. In: Annual Review of Statistics and Its Application 1.1 (2014), pp. 125–151.</p>
</div>
<div class="citation" id="fissler2022" role="doc-biblioentry">
<span class="label"><span class="fn-bracket">[</span><a role="doc-backlink" href="#id5">Fissler2022</a><span class="fn-bracket">]</span></span>
<p>T. Fissler, C. Lorentzen and M. Mayer. <a class="reference external" href="https://arxiv.org/abs/2202.12780">Model
Comparison and Calibration Assessment: User Guide for Consistent Scoring
Functions in Machine Learning and Actuarial Practice.</a></p>
</div>
</div>
</section>
<section id="scoring-api-overview">
<span id="id6"></span><h2><span class="section-number">3.4.2. </span>Scoring API overview<a class="headerlink" href="#scoring-api-overview" title="Link to this heading">#</a></h2>
<p>There are 3 different APIs for evaluating the quality of a model’s
predictions:</p>
<ul class="simple">
<li><p><strong>Estimator score method</strong>: Estimators have a <code class="docutils literal notranslate"><span class="pre">score</span></code> method providing a
default evaluation criterion for the problem they are designed to solve.
Most commonly this is <a class="reference internal" href="#accuracy-score"><span class="std std-ref">accuracy</span></a> for classifiers and the
<a class="reference internal" href="#r2-score"><span class="std std-ref">coefficient of determination</span></a> (<span class="math notranslate nohighlight">\(R^2\)</span>) for regressors.
Details for each estimator can be found in its documentation.</p></li>
<li><p><strong>Scoring parameter</strong>: Model-evaluation tools that use
<a class="reference internal" href="cross_validation.html#cross-validation"><span class="std std-ref">cross-validation</span></a> (such as
<a class="reference internal" href="generated/sklearn.model_selection.GridSearchCV.html#sklearn.model_selection.GridSearchCV" title="sklearn.model_selection.GridSearchCV"><code class="xref py py-class docutils literal notranslate"><span class="pre">model_selection.GridSearchCV</span></code></a>, <a class="reference internal" href="generated/sklearn.model_selection.validation_curve.html#sklearn.model_selection.validation_curve" title="sklearn.model_selection.validation_curve"><code class="xref py py-func docutils literal notranslate"><span class="pre">model_selection.validation_curve</span></code></a> and
<a class="reference internal" href="generated/sklearn.linear_model.LogisticRegressionCV.html#sklearn.linear_model.LogisticRegressionCV" title="sklearn.linear_model.LogisticRegressionCV"><code class="xref py py-class docutils literal notranslate"><span class="pre">linear_model.LogisticRegressionCV</span></code></a>) rely on an internal <em>scoring</em> strategy.
This can be specified using the <code class="docutils literal notranslate"><span class="pre">scoring</span></code> parameter of that tool and is discussed
in the section <a class="reference internal" href="#scoring-parameter"><span class="std std-ref">The scoring parameter: defining model evaluation rules</span></a>.</p></li>
<li><p><strong>Metric functions</strong>: The <a class="reference internal" href="../api/sklearn.metrics.html#module-sklearn.metrics" title="sklearn.metrics"><code class="xref py py-mod docutils literal notranslate"><span class="pre">sklearn.metrics</span></code></a> module implements functions
assessing prediction error for specific purposes. These metrics are detailed
in sections on <a class="reference internal" href="#classification-metrics"><span class="std std-ref">Classification metrics</span></a>,
<a class="reference internal" href="#multilabel-ranking-metrics"><span class="std std-ref">Multilabel ranking metrics</span></a>, <a class="reference internal" href="#regression-metrics"><span class="std std-ref">Regression metrics</span></a> and
<a class="reference internal" href="#clustering-metrics"><span class="std std-ref">Clustering metrics</span></a>.</p></li>
</ul>
<p>Finally, <a class="reference internal" href="#dummy-estimators"><span class="std std-ref">Dummy estimators</span></a> are useful to get a baseline
value of those metrics for random predictions.</p>
<div class="admonition seealso">
<p class="admonition-title">See also</p>
<p>For “pairwise” metrics, between <em>samples</em> and not estimators or
predictions, see the <a class="reference internal" href="metrics.html#metrics"><span class="std std-ref">Pairwise metrics, Affinities and Kernels</span></a> section.</p>
</div>
</section>
<section id="the-scoring-parameter-defining-model-evaluation-rules">
<span id="scoring-parameter"></span><h2><span class="section-number">3.4.3. </span>The <code class="docutils literal notranslate"><span class="pre">scoring</span></code> parameter: defining model evaluation rules<a class="headerlink" href="#the-scoring-parameter-defining-model-evaluation-rules" title="Link to this heading">#</a></h2>
<p>Model selection and evaluation tools that internally use
<a class="reference internal" href="cross_validation.html#cross-validation"><span class="std std-ref">cross-validation</span></a> (such as
<a class="reference internal" href="generated/sklearn.model_selection.GridSearchCV.html#sklearn.model_selection.GridSearchCV" title="sklearn.model_selection.GridSearchCV"><code class="xref py py-class docutils literal notranslate"><span class="pre">model_selection.GridSearchCV</span></code></a>, <a class="reference internal" href="generated/sklearn.model_selection.validation_curve.html#sklearn.model_selection.validation_curve" title="sklearn.model_selection.validation_curve"><code class="xref py py-func docutils literal notranslate"><span class="pre">model_selection.validation_curve</span></code></a> and
<a class="reference internal" href="generated/sklearn.linear_model.LogisticRegressionCV.html#sklearn.linear_model.LogisticRegressionCV" title="sklearn.linear_model.LogisticRegressionCV"><code class="xref py py-class docutils literal notranslate"><span class="pre">linear_model.LogisticRegressionCV</span></code></a>) take a <code class="docutils literal notranslate"><span class="pre">scoring</span></code> parameter that
controls what metric they apply to the estimators evaluated.</p>
<p>They can be specified in several ways:</p>
<ul class="simple">
<li><p><code class="docutils literal notranslate"><span class="pre">None</span></code>: the estimator’s default evaluation criterion (i.e., the metric used in the
estimator’s <code class="docutils literal notranslate"><span class="pre">score</span></code> method) is used.</p></li>
<li><p><a class="reference internal" href="#scoring-string-names"><span class="std std-ref">String name</span></a>: common metrics can be passed via a string
name.</p></li>
<li><p><a class="reference internal" href="#scoring-callable"><span class="std std-ref">Callable</span></a>: more complex metrics can be passed via a custom
metric callable (e.g., function).</p></li>
</ul>
<p>Some tools do also accept multiple metric evaluation. See <a class="reference internal" href="#multimetric-scoring"><span class="std std-ref">Using multiple metric evaluation</span></a>
for details.</p>
<section id="string-name-scorers">
<span id="scoring-string-names"></span><h3><span class="section-number">3.4.3.1. </span>String name scorers<a class="headerlink" href="#string-name-scorers" title="Link to this heading">#</a></h3>
<p>For the most common use cases, you can designate a scorer object with the
<code class="docutils literal notranslate"><span class="pre">scoring</span></code> parameter via a string name; the table below shows all possible values.
All scorer objects follow the convention that <strong>higher return values are better
than lower return values</strong>. Thus metrics which measure the distance between
the model and the data, like <a class="reference internal" href="generated/sklearn.metrics.mean_squared_error.html#sklearn.metrics.mean_squared_error" title="sklearn.metrics.mean_squared_error"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.mean_squared_error</span></code></a>, are
available as ‘neg_mean_squared_error’ which return the negated value
of the metric.</p>
<div class="pst-scrollable-table-container"><table class="table">
<thead>
<tr class="row-odd"><th class="head"><p>Scoring string name</p></th>
<th class="head"><p>Function</p></th>
<th class="head"><p>Comment</p></th>
</tr>
</thead>
<tbody>
<tr class="row-even"><td><p><strong>Classification</strong></p></td>
<td></td>
<td></td>
</tr>
<tr class="row-odd"><td><p>‘accuracy’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.accuracy_score.html#sklearn.metrics.accuracy_score" title="sklearn.metrics.accuracy_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.accuracy_score</span></code></a></p></td>
<td></td>
</tr>
<tr class="row-even"><td><p>‘balanced_accuracy’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.balanced_accuracy_score.html#sklearn.metrics.balanced_accuracy_score" title="sklearn.metrics.balanced_accuracy_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.balanced_accuracy_score</span></code></a></p></td>
<td></td>
</tr>
<tr class="row-odd"><td><p>‘top_k_accuracy’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.top_k_accuracy_score.html#sklearn.metrics.top_k_accuracy_score" title="sklearn.metrics.top_k_accuracy_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.top_k_accuracy_score</span></code></a></p></td>
<td></td>
</tr>
<tr class="row-even"><td><p>‘average_precision’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.average_precision_score.html#sklearn.metrics.average_precision_score" title="sklearn.metrics.average_precision_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.average_precision_score</span></code></a></p></td>
<td></td>
</tr>
<tr class="row-odd"><td><p>‘neg_brier_score’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.brier_score_loss.html#sklearn.metrics.brier_score_loss" title="sklearn.metrics.brier_score_loss"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.brier_score_loss</span></code></a></p></td>
<td></td>
</tr>
<tr class="row-even"><td><p>‘f1’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.f1_score.html#sklearn.metrics.f1_score" title="sklearn.metrics.f1_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.f1_score</span></code></a></p></td>
<td><p>for binary targets</p></td>
</tr>
<tr class="row-odd"><td><p>‘f1_micro’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.f1_score.html#sklearn.metrics.f1_score" title="sklearn.metrics.f1_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.f1_score</span></code></a></p></td>
<td><p>micro-averaged</p></td>
</tr>
<tr class="row-even"><td><p>‘f1_macro’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.f1_score.html#sklearn.metrics.f1_score" title="sklearn.metrics.f1_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.f1_score</span></code></a></p></td>
<td><p>macro-averaged</p></td>
</tr>
<tr class="row-odd"><td><p>‘f1_weighted’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.f1_score.html#sklearn.metrics.f1_score" title="sklearn.metrics.f1_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.f1_score</span></code></a></p></td>
<td><p>weighted average</p></td>
</tr>
<tr class="row-even"><td><p>‘f1_samples’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.f1_score.html#sklearn.metrics.f1_score" title="sklearn.metrics.f1_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.f1_score</span></code></a></p></td>
<td><p>by multilabel sample</p></td>
</tr>
<tr class="row-odd"><td><p>‘neg_log_loss’</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.log_loss.html#sklearn.metrics.log_loss" title="sklearn.metrics.log_loss"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.log_loss</span></code></a></p></td>
<td><p>requires <code class="docutils literal notranslate"><span class="pre">predict_proba</span></code> support</p></td>
</tr>
<tr class="row-even"><td><p>‘precision’ etc.</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.precision_score.html#sklearn.metrics.precision_score" title="sklearn.metrics.precision_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.precision_score</span></code></a></p></td>
<td><p>suffixes apply as with ‘f1’</p></td>
</tr>
<tr class="row-odd"><td><p>‘recall’ etc.</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.recall_score.html#sklearn.metrics.recall_score" title="sklearn.metrics.recall_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.recall_score</span></code></a></p></td>
<td><p>suffixes apply as with ‘f1’</p></td>
</tr>
<tr class="row-even"><td><p>‘jaccard’ etc.</p></td>
<td><p><a class="reference internal" href="generated/sklearn.metrics.jaccard_score.html#sklearn.metrics.jaccard_score" title="sklearn.metrics.jaccard_score"><code class="xref py py-func docutils literal notranslate"><span class="pre">metrics.jaccard_score</span></code></a></p></td>
<td><p>suffixes apply as with ‘f1’</p></td>