Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
11
0

Year Published

2013
2013
2014
2014

Publication Types

Select...
1
1

Relationship

2
0

Authors

Journals

##### Cited by 14 publications
(11 citation statements)
##### References 18 publications
(11 reference statements)
0
11
0
Order By: Relevance
“…The second component is the local proportion that has the same mathematical form as global proportion, however it is calculated taking into account only the examples from neighborhood. The neighborhood is composed of all similar examples, i.e., examples, which share the same description on nominal features and have the value in the same interval on numerical ones (i.e., local discretization) [3]. We consider neighborhoods which are constructed on random subsets of the original feature set (see [2]).…”
Section: New Proposalsmentioning
confidence: 99%
“…The second component is the local proportion that has the same mathematical form as global proportion, however it is calculated taking into account only the examples from neighborhood. The neighborhood is composed of all similar examples, i.e., examples, which share the same description on nominal features and have the value in the same interval on numerical ones (i.e., local discretization) [3]. We consider neighborhoods which are constructed on random subsets of the original feature set (see [2]).…”
Section: New Proposalsmentioning
confidence: 99%
“…The samples of objects used in the induction process were controlled by consistency measures. The approach applied to this end is called VC-bagging [1,2]. It extends the standard bagging scheme proposed by Breiman [6].…”
Section: Methodsmentioning
confidence: 99%
“…We take, moreover, into account the postulate saying that base classifiers used in bagging are expected to have sufficiently high predictive accuracy apart from being diversified [7]. As we have shown in previous studies [1,2], this requirement can be satisfied by privileging consistent objects when generating bootstrap samples. Presence of inconsistent objects in bootstrap samples may lead to overfitting of the base classifies, which decreases their classification accuracy.…”
Section: Methodsmentioning
confidence: 99%
See 2 more Smart Citations