scikit-learn
diff --git a/‎doc/modules/model_evaluation.rst
Copy file name to clipboardExpand all lines: doc/modules/model_evaluation.rst
+8-41Lines changed: 8 additions & 41 deletions b/‎doc/modules/model_evaluation.rst
Copy file name to clipboardExpand all lines: doc/modules/model_evaluation.rst
+8-41Lines changed: 8 additions & 41 deletions
@@ -580,9 +580,7 @@ or *informedness*.
     * Balanced Accuracy as described in [Urbanowicz2015]_: the average of sensitivity and specificity
       is computed for each class and then averaged over total number of classes.
 
-|details-start|
-**References**
-|details-split|
+.. topic:: References:
 
   .. [Guyon2015] I. Guyon, K. Bennett, G. Cawley, H.J. Escalante, S. Escalera, T.K. Ho, N. Macià,
      B. Ray, M. Saeed, A.R. Statnikov, E. Viegas, `Design of the 2015 ChaLearn AutoML Challenge
@@ -599,8 +597,6 @@ or *informedness*.
       and evaluation of a scalable learning classifier
       system <10.1007/s12065-015-0128-8>`, Evol. Intel. (2015) 8: 89.
 
-|details-end|
-
 .. _cohen_kappa:
 
 Cohen's kappa
@@ -853,9 +849,7 @@ precision-recall curve as follows.
     for an example of :func:`precision_recall_curve` usage to evaluate
     classifier output quality.
 
-|details-start|
-**References**
-|details-split|
+.. topic:: References:
 
   .. [Manning2008] C.D. Manning, P. Raghavan, H. Schütze, `Introduction to Information Retrieval
      <https://nlp.stanford.edu/IR-book/html/htmledition/evaluation-of-ranked-retrieval-results-1.html>`_,
@@ -871,9 +865,6 @@ precision-recall curve as follows.
      <https://papers.nips.cc/paper/5867-precision-recall-gain-curves-pr-analysis-done-right.pdf>`_,
      NIPS 2015.
 
-|details-end|
-
-
 Binary classification
 ^^^^^^^^^^^^^^^^^^^^^
 
@@ -1470,9 +1461,7 @@ correspond to the probability estimates that a sample belongs to a particular
 class. The OvO and OvR algorithms support weighting uniformly
 (``average='macro'``) and by prevalence (``average='weighted'``).
 
-|details-start|
-**One-vs-one Algorithm**
-|details-split|
+.. topic:: One-vs-one Algorithm:
 
 Computes the average AUC of all possible pairwise
 combinations of classes. [HT2001]_ defines a multiclass AUC metric weighted
@@ -1503,11 +1492,7 @@ the keyword argument ``multiclass`` to ``'ovo'`` and ``average`` to
 ``'weighted'``. The ``'weighted'`` option returns a prevalence-weighted average
 as described in [FC2009]_.
 
-|details-end|
-
-|details-start|
-**One-vs-rest Algorithm**
-|details-split|
+.. topic:: One-vs-rest Algorithm:
 
 Computes the AUC of each class against the rest
 [PD2000]_. The algorithm is functionally the same as the multilabel case. To
@@ -1528,8 +1513,6 @@ the :ref:`iris_dataset`:
    :scale: 75
    :align: center
 
-|details-end|
-
 .. _roc_auc_multilabel:
 
 Multi-label case
@@ -1572,9 +1555,7 @@ And the decision values do not require such processing.
     for an example of using ROC to
     model species distribution.
 
-|details-start|
-**References**
-|details-split|
+.. topic:: References:
 
     .. [HT2001] Hand, D.J. and Till, R.J., (2001). `A simple generalisation
        of the area under the ROC curve for multiple class classification problems.
@@ -1601,8 +1582,6 @@ And the decision values do not require such processing.
        In Data Mining, 2001.
        Proceedings IEEE International Conference, pp. 131-138.
 
-|details-end|
-
 .. _det_curve:
 
 Detection error tradeoff (DET)
@@ -1643,9 +1622,7 @@ same classification task:
     for an example comparison between receiver operating characteristic (ROC)
     curves and Detection error tradeoff (DET) curves.
 
-|details-start|
-**Properties**
-|details-split|
+.. topic:: Properties:
 
 * DET curves form a linear curve in normal deviate scale if the detection
   scores are normally (or close-to normally) distributed.
@@ -1661,8 +1638,6 @@ same classification task:
   of perfection for DET curves is the origin (in contrast to the top left
   corner for ROC curves).
 
-|details-end|
-
 |details-start|
 **Applications and limitations**
 |details-split|
@@ -1680,9 +1655,7 @@ better suited.
 
 |details-end|
 
-|details-start|
-**References**
-|details-split|
+.. topic:: References:
 
   .. [WikipediaDET2017] Wikipedia contributors. Detection error tradeoff.
      Wikipedia, The Free Encyclopedia. September 4, 2017, 23:33 UTC.
@@ -1701,8 +1674,6 @@ better suited.
      Speech and Signal Processing - ICASSP '07, Honolulu,
      HI, 2007, pp. IV-229-IV-232.
 
-|details-end|
-
 .. _zero_one_loss:
 
 Zero one loss
@@ -1815,9 +1786,7 @@ the same does a lower Brier score loss always mean better calibration"
     for an example of Brier score loss usage to perform probability
     calibration of classifiers.
 
-|details-start|
-**References**
-|details-split|
+.. topic:: References:
 
   .. [Brier1950] G. Brier, `Verification of forecasts expressed in terms of
     probability
@@ -1834,8 +1803,6 @@ the same does a lower Brier score loss always mean better calibration"
     and probability estimation." <https://drops.dagstuhl.de/opus/volltexte/2008/1382/>`_
     Dagstuhl Seminar Proceedings. Schloss Dagstuhl-Leibniz-Zentrum fr Informatik (2008).
 
-|details-end|
-
 .. _class_likelihood_ratios:
 
 Class likelihood ratios