http://debianws.lexgopc.com/wiki143/index.php?action=history&feed=atom&title=CN2_algorithm 2026-10-26T12:47:14Z Revision history for this page on the wiki MediaWiki 1.43.1 http://debianws.lexgopc.com/wiki143/index.php?title=CN2_algorithm&diff=4063606&oldid=prev 2025-06-26T18:55:13Z

<p><span class="autocomment">External links</span></p> <p><b>New page</b></p><div>{{Refimprove|date=September 2012}}<br /> The &#039;&#039;&#039;CN2 induction algorithm&#039;&#039;&#039; is a [[Algorithmic learning theory|learning algorithm]] for [[rule induction]].&lt;ref&gt;Clark, P. and Niblett, T (1989) The CN2 induction algorithm. Machine Learning 3(4):261-283.&lt;/ref&gt; It is designed to work even when the training data is imperfect. It is based on ideas from the [[AQ algorithm]] and the [[ID3 algorithm]]. As a consequence it creates a rule set like that created by AQ but is able to handle noisy data like ID3.<br /> <br /> ==Description of algorithm==<br /> The algorithm must be given a set of examples, TrainingSet, which have already been classified in order to generate a list of classification rules. A set of conditions, SimpleConditionSet, which can be applied, alone or in combination, to any set of examples is predefined to be used for the classification.<br /> <br /> routine CN2(TrainingSet)<br /> let the ClassificationRuleList be empty<br /> repeat<br /> let the BestConditionExpression be Find_BestConditionExpression(TrainingSet)<br /> if the BestConditionExpression is not nil<br /> then<br /> let the TrainingSubset be the examples covered by the BestConditionExpression<br /> remove from the TrainingSet the examples in the TrainingSubset<br /> let the MostCommonClass be the most common class of examples in the TrainingSubset<br /> append to the ClassificationRuleList the rule<br /> &#039;if &#039; the BestConditionExpression &#039; then the class is &#039; the MostCommonClass<br /> until the TrainingSet is empty or the BestConditionExpression is nil<br /> return the ClassificationRuleList<br /> <br /> routine Find_BestConditionExpression(TrainingSet)<br /> let the ConditionalExpressionSet be empty<br /> let the BestConditionExpression be nil<br /> repeat<br /> let the TrialConditionalExpressionSet be the set of conditional expressions,<br /> {x and y where x belongs to the ConditionalExpressionSet and y belongs to the SimpleConditionSet}.<br /> remove all formulae in the TrialConditionalExpressionSet that are either in the ConditionalExpressionSet (i.e.,<br /> the unspecialized ones) or null (e.g., big = y and big = n)<br /> for every expression, F, in the TrialConditionalExpressionSet<br /> if<br /> F is statistically significant<br /> and F is better than the BestConditionExpression<br /> by user-defined criteria when tested on the TrainingSet<br /> then<br /> replace the current value of the BestConditionExpression by F<br /> while the number of expressions in the TrialConditionalExpressionSet &gt; user-defined maximum<br /> remove the worst expression from the TrialConditionalExpressionSet<br /> let the ConditionalExpressionSet be the TrialConditionalExpressionSet<br /> until the ConditionalExpressionSet is empty<br /> return the BestConditionExpression<br /> <br /> ==References==<br /> &lt;references /&gt;<br /> <br /> ==External links==<br /> * [https://doi.org/10.1023%2FA%3A1022641700528 CN2 Algorithm Description]<br /> <br /> [[Category:Machine learning algorithms]]<br /> <br /> <br /> {{algorithm-stub}}<br /> {{machine-learning-stub}}</div>

imported>Fadesga