source: orange/docs/reference/rst/Orange.evaluation.rst @ 10889:a63021ee6a68

Revision 10889:a63021ee6a68, 1.3 KB checked in by Matija Polajnar <matija.polajnar@…>, 23 months ago (diff)

Move of reliability into an add-on.

Line 
1#############################
2Evaluation (``evaluation``)
3#############################
4
5Evaluation of prediction modules is split into two parts. Module
6:obj:`Orange.evaluation.testing` contains procedures that sample data,
7train learning algorithms and test models. All procedures return
8results as an instance of
9:obj:`~Orange.evaluation.testing.ExperimentResults` that is described
10below. Module :obj:`Orange.evaluation.scoring` uses such data to
11compute various performance scores like classification accuracy and
12AUC.
13
14There is a third module available as an add-on, which is unrelated to this
15scheme,:obj:`Orange.evaluation.reliability`, that assesses the reliability
16of individual predictions.
17
18.. toctree::
19   :maxdepth: 1
20
21   Orange.evaluation.testing
22   Orange.evaluation.scoring
23
24Classes for storing the experimental results
25--------------------------------------------
26
27
28The following two classes are used for storing the results of experiments by :obj:`Orange.evaluation.testing` and computing of scores by :obj:`Orange.evaluation.scoring`. Instances of this class seldom need to be constructed and used outside of these two modules.
29
30.. py:currentmodule:: Orange.evaluation.testing
31
32.. autoclass:: ExperimentResults(iterations, classifier_names, class_values=None, weights=None, base_class=-1)
33    :members:
34
35.. autoclass:: TestedExample
36    :members:
Note: See TracBrowser for help on using the repository browser.