Measuring relevance between discrete and continuous features based on neighborhood mutual information
Citations
178 citations
158 citations
Cites background or methods from "Measuring relevance between discret..."
...where [xi ]R1 is the successor neighborhood of xi with respect to R1 (see [16], [17])....
[...]
...According to neighborhood entropy [16], [17]...
[...]
...neighborhood rough set-based algorithm (NRS) [15], neighborhood entropy-based algorithm (NEIEN) [16], fuzzy information entropy-based algorithm (FINEN) [17], [58], and fuzzy rough dependence constructed by intersection operations of...
[...]
...To present the selected feature subset of a data set, in the following we employ the NEIEN, FINEN, and HANDI algorithms to reduce the entire data set based on the parameters where the classification accuracies were obtained in the above experiments....
[...]
...The complexity of HANDI is less than the NEIEN, FINEN, and FRSINT algorithms....
[...]
147 citations
Cites background or methods from "Measuring relevance between discret..."
...Thus, the Nemenyi tests demonstrate that NSI is significantly better than NRE and NRS at α = 0.1, respectively....
[...]
...[54] defined a feature relevance measure to characterize the classification ability of feature subsets....
[...]
...Three excellent algorithms, including neighborhood entropy (NRE) [54], NRS [14] and neighborhood discrimination index (NDI) [56], are selected and used to compare the proposed method....
[...]
...For the 3NN classifier, it is easily observed from Table IX that the distances between NSI to NRE, NRS, and NDI are all greater than 1.2612....
[...]
...According to Table VIII, the distances between average orderings of NSI to NRE and NRS are greater than 1.2612 for SVM....
[...]
120 citations
117 citations
References
65,425 citations
"Measuring relevance between discret..." refers background in this paper
...Shannon’s entropy, first introduced in 1948 (Shannon, 1948), is a measure of uncertainty of random variables....
[...]
21,694 citations
17,177 citations
16,974 citations
"Measuring relevance between discret..." refers methods in this paper
...In decision tree construction, indexes such as Gini, towing, deviance and mutual information were introduced to compute the relevance between inputs and output, thus guilding the algorithms to select an informative feature to split samples (Breiman, 1993; Quinlan, 1986, 1993)....
[...]
14,825 citations