Matches in SemOpenAlex for { <https://semopenalex.org/work/W2945235690> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W2945235690 endingPage "274" @default.
- W2945235690 startingPage "255" @default.
- W2945235690 abstract "In recent years feature selection methods based on mutual information have attracted a significant attention. Most of the proposed methods are based on sequential forward search which add at each step a feature which is most relevant in explaining a class variable when considered together with the already chosen features. Such procedures produce ranking of features ordered according to their relevance. However significant limitation of all existing methods is lack of stopping rules which separate relevant features placed on the top of the ranking list from irrelevant ones. Finding an appropriate stopping rule is particularly important in domains where one wants to precisely determine the set of features affecting the class variable and discard the irrelevant ones (e.g. in genome-wide association studies the goal is to precisely determine mutations in DNA affecting the disease). In this work we propose stopping rules which are based on distribution of approximation of conditional mutual information given that all relevant features have been already selected. We show that the distribution is approximately chi square with appropriate number of degrees of freedom provided features are discretized into moderate number of bins. The proposed stopping rules are based on quantiles of the distribution and related p-values which are compared with thresholds used in multiple hypothesis testing. Importantly the proposed methods do not require additional validation data and are independent from the classifier. The extensive simulation experiments indicate that the rules separate relevant features from the irrelevant ones. We show experimentally that Positive Selection Rate (fraction of features correctly selected as relevant with respect to all relevant features) approaches 1, when sample size increases. At the same time, False Discovery Rate (fraction of irrelevant features selected with respect to all selected features) is controlled. The experiments on 17 benchmark datasets indicate that the classification models, built on features selected by the proposed methods, in 13 cases achieve significantly higher accuracy than the models based on all available features." @default.
- W2945235690 created "2019-05-29" @default.
- W2945235690 creator A5037622299 @default.
- W2945235690 creator A5084772706 @default.
- W2945235690 date "2019-09-01" @default.
- W2945235690 modified "2023-09-23" @default.
- W2945235690 title "Stopping rules for mutual information-based feature selection" @default.
- W2945235690 cites W1822695366 @default.
- W2945235690 cites W1965138864 @default.
- W2945235690 cites W1965680102 @default.
- W2945235690 cites W1972401983 @default.
- W2945235690 cites W2044170702 @default.
- W2945235690 cites W2063249739 @default.
- W2945235690 cites W2098185113 @default.
- W2945235690 cites W2114771311 @default.
- W2945235690 cites W2118561568 @default.
- W2945235690 cites W2149772057 @default.
- W2945235690 cites W2154053567 @default.
- W2945235690 cites W2165580920 @default.
- W2945235690 cites W2177148767 @default.
- W2945235690 cites W2290195878 @default.
- W2945235690 cites W2522585561 @default.
- W2945235690 cites W2570428969 @default.
- W2945235690 cites W2777199146 @default.
- W2945235690 cites W2785531387 @default.
- W2945235690 cites W2802045099 @default.
- W2945235690 cites W2885496398 @default.
- W2945235690 cites W2887002214 @default.
- W2945235690 doi "https://doi.org/10.1016/j.neucom.2019.05.048" @default.
- W2945235690 hasPublicationYear "2019" @default.
- W2945235690 type Work @default.
- W2945235690 sameAs 2945235690 @default.
- W2945235690 citedByCount "16" @default.
- W2945235690 countsByYear W29452356902019 @default.
- W2945235690 countsByYear W29452356902020 @default.
- W2945235690 countsByYear W29452356902021 @default.
- W2945235690 countsByYear W29452356902022 @default.
- W2945235690 countsByYear W29452356902023 @default.
- W2945235690 crossrefType "journal-article" @default.
- W2945235690 hasAuthorship W2945235690A5037622299 @default.
- W2945235690 hasAuthorship W2945235690A5084772706 @default.
- W2945235690 hasConcept C105795698 @default.
- W2945235690 hasConcept C118671147 @default.
- W2945235690 hasConcept C119857082 @default.
- W2945235690 hasConcept C124101348 @default.
- W2945235690 hasConcept C138885662 @default.
- W2945235690 hasConcept C148483581 @default.
- W2945235690 hasConcept C152139883 @default.
- W2945235690 hasConcept C153180895 @default.
- W2945235690 hasConcept C154945302 @default.
- W2945235690 hasConcept C189430467 @default.
- W2945235690 hasConcept C2776401178 @default.
- W2945235690 hasConcept C33923547 @default.
- W2945235690 hasConcept C41008148 @default.
- W2945235690 hasConcept C41895202 @default.
- W2945235690 hasConcept C50644808 @default.
- W2945235690 hasConcept C5465570 @default.
- W2945235690 hasConcept C95623464 @default.
- W2945235690 hasConceptScore W2945235690C105795698 @default.
- W2945235690 hasConceptScore W2945235690C118671147 @default.
- W2945235690 hasConceptScore W2945235690C119857082 @default.
- W2945235690 hasConceptScore W2945235690C124101348 @default.
- W2945235690 hasConceptScore W2945235690C138885662 @default.
- W2945235690 hasConceptScore W2945235690C148483581 @default.
- W2945235690 hasConceptScore W2945235690C152139883 @default.
- W2945235690 hasConceptScore W2945235690C153180895 @default.
- W2945235690 hasConceptScore W2945235690C154945302 @default.
- W2945235690 hasConceptScore W2945235690C189430467 @default.
- W2945235690 hasConceptScore W2945235690C2776401178 @default.
- W2945235690 hasConceptScore W2945235690C33923547 @default.
- W2945235690 hasConceptScore W2945235690C41008148 @default.
- W2945235690 hasConceptScore W2945235690C41895202 @default.
- W2945235690 hasConceptScore W2945235690C50644808 @default.
- W2945235690 hasConceptScore W2945235690C5465570 @default.
- W2945235690 hasConceptScore W2945235690C95623464 @default.
- W2945235690 hasLocation W29452356901 @default.
- W2945235690 hasOpenAccess W2945235690 @default.
- W2945235690 hasPrimaryLocation W29452356901 @default.
- W2945235690 hasRelatedWork W2018051788 @default.
- W2945235690 hasRelatedWork W2053982601 @default.
- W2945235690 hasRelatedWork W2884983377 @default.
- W2945235690 hasRelatedWork W2922409390 @default.
- W2945235690 hasRelatedWork W2946101226 @default.
- W2945235690 hasRelatedWork W2952668426 @default.
- W2945235690 hasRelatedWork W3136130306 @default.
- W2945235690 hasRelatedWork W3200179079 @default.
- W2945235690 hasRelatedWork W4224118342 @default.
- W2945235690 hasRelatedWork W1933296232 @default.
- W2945235690 hasVolume "358" @default.
- W2945235690 isParatext "false" @default.
- W2945235690 isRetracted "false" @default.
- W2945235690 magId "2945235690" @default.
- W2945235690 workType "article" @default.