Changeset 42:75bf74617e81 in orange-reliability for docs/rst/Orange.evaluation.reliability.rst


Ignore:
Timestamp:
10/03/13 16:29:59 (7 months ago)
Author:
markotoplak
Branch:
default
Message:

Updates to the documentation.

File:
1 edited

Legend:

Unmodified
Added
Removed
  • docs/rst/Orange.evaluation.reliability.rst

    r41 r42  
    1515 
    1616Reliability assessment aims to predict reliabilities of individual 
    17 predictions.  
    18  
    19 Most of implemented algorithms for regression described in 
    20 "Comparison of approaches for estimating reliability of individual 
    21 regression predictions, Zoran Bosnić, 2008" for regression and in 
    22 in "Evaluating Reliability of Single 
    23 Classifications of Neural Networks, Darko Pevec, 2011" for classification. 
     17predictions. Most of implemented algorithms for regression described in 
     18[Bosnic2008]_ and in [Pevec2011]_ for classification. 
    2419 
    2520We can use reliability estimation with any Orange learners. The following example: 
    2621 
    2722 * Constructs reliability estimators (implemented in this module), 
    28  * Combines a regular learner. 
    29    (:class:`~Orange.classification.knn.kNNLearner` in this case) with 
    30    reliability estimators. 
     23 * The :obj:`Learner` wrapper combines a regular learner, here a :obj:`~Orange.classification.knn.kNNLearner`, with reliability estimators. 
    3124 * Obtains prediction probabilities from the constructed classifier 
    3225   (:obj:`Orange.classification.Classifier.GetBoth` option). The resulting 
    33    probabilities have and additional attribute, :obj:`reliability_estimate` 
    34    attribute, :class:`Orange.evaluation.reliability.Estimate`. 
     26   probabilities have an additional attribute, :obj:`reliability_estimate`, 
     27   that contains a list of :class:`Orange.evaluation.reliability.Estimate`. 
    3528 
    3629.. literalinclude:: code/reliability-basic.py 
     
    4336    :lines: 7- 
    4437 
     38Reliability estimation wrappers 
     39=============================== 
     40 
     41.. autoclass:: Learner 
     42   :members: __call__ 
     43 
     44.. autoclass:: Classifier 
     45   :members: __call__ 
     46 
     47 
    4548Reliability Methods 
    4649=================== 
    4750 
    48 For regression, you can use all the described measures except :math:`O_{ref}`. Classification is 
     51All measures except :math:`O_{ref}` work with regression. Classification is 
    4952supported by BAGV, LCV, CNK and DENS, :math:`O_{ref}`. 
    5053 
     
    9295 
    9396Stacked generalization (Stacking) 
    94 ------------------------------- 
     97--------------------------------- 
    9598 
    9699.. autoclass:: Stacking 
     
    100103 
    101104.. autoclass:: ReferenceExpectedError 
    102  
    103 Reliability estimation wrappers 
    104 =============================== 
    105  
    106 .. autoclass:: Learner(box_learner, name="Reliability estimation", estimators=[SensitivityAnalysis(), LocalCrossValidation(), BaggingVarianceCNeighbours(), Mahalanobis(), MahalanobisToCenter()], **kwds) 
    107     :members: 
    108  
    109 .. autoclass:: Classifier 
    110     :members: 
    111105 
    112106Reliability estimation results 
     
    167161========== 
    168162 
    169 Bosnić, Z., Kononenko, I. (2007) `Estimation of individual prediction 
    170 reliability using local sensitivity analysis. <http://www.springerlink 
    171 .com/content/e27p2584387532g8/>`_ *Applied Intelligence* 29(3), pp. 187-203. 
     163.. [Bosnic2007]  Bosnić, Z., Kononenko, I. (2007) `Estimation of individual prediction reliability using local sensitivity analysis. <http://www.springerlink.com/content/e27p2584387532g8/>`_ *Applied Intelligence* 29(3), pp. 187-203. 
    172164 
    173 Bosnić, Z., Kononenko, I. (2008) `Comparison of approaches for estimating 
    174 reliability of individual regression predictions. <http://www.sciencedirect 
    175 .com/science/article/pii/S0169023X08001080>`_ *Data & Knowledge Engineering* 
    176 67(3), pp. 504-516. 
     165.. [Bosnic2008] Bosnić, Z., Kononenko, I. (2008) `Comparison of approaches for estimating reliability of individual regression predictions. <http://www.sciencedirect .com/science/article/pii/S0169023X08001080>`_ *Data & Knowledge Engineering* 67(3), pp. 504-516. 
    177166 
    178 Bosnić, Z., Kononenko, I. (2010) `Automatic selection of reliability estimates 
    179 for individual regression predictions. <http://journals.cambridge 
    180 .org/abstract_S0269888909990154>`_ *The Knowledge Engineering Review* 25(1), 
    181 pp. 27-47. 
     167.. [Bosnic2010] Bosnić, Z., Kononenko, I. (2010) `Automatic selection of reliability estimates for individual regression predictions. <http://journals.cambridge .org/abstract_S0269888909990154>`_ *The Knowledge Engineering Review* 25(1), pp. 27-47. 
    182168 
    183 Pevec, D., Štrumbelj, E., Kononenko, I. (2011) `Evaluating Reliability of 
    184 Single Classifications of Neural Networks. <http://www.springerlink.com 
    185 /content/48u881761h127r33/export-citation/>`_ *Adaptive and Natural Computing 
    186 Algorithms*, 2011, pp. 22-30. 
     169.. [Pevec2011] Pevec, D., Štrumbelj, E., Kononenko, I. (2011) `Evaluating Reliability of Single Classifications of Neural Networks. <http://www.springerlink.com /content/48u881761h127r33/export-citation/>`_ *Adaptive and Natural Computing Algorithms*, 2011, pp. 22-30. 
Note: See TracChangeset for help on using the changeset viewer.