A functional neural fuzzy network for classification applications

A hierarchical approach to multi-class fuzzy classifiers

“…We applied our method to this dataset and we compare our results with those achieved by some authors in the literature Ait Kbir et al, 2000;Ishibuchi et al, 2005a;Lee et al, 2001;Sanz et al, 2011;Setnes & Roubos, 2000;Tung et al, 2012;Wang & Lee, 2002;Wu et al, 2011). Table 2 shows the results achieved by the aforementioned authors on the Iris dataset along with our results (first row of Table 2).…”

“…The achieved results outperform those obtained by other methods found in the literature. 3.6 13 -91.22 HFP (Guillaume & Charnomordic, 2004) 6.8 4 -89.2 HGBML (Ishibuchi et al, 2005a) 10 13 -95.06 ''RW'' (Ishibuchi & Yamamoto, 2005b) 15 13 -95.51 HCGA 4.9 --95.66 FH-GBML -13 -93.79 ''TSK-GA'' (Setnes & Roubos, 2000) 3 9 21 98.3 SANFIS (Wang & Lee, 2002) 3 13 34 99.4 FNFN (Wu et al, 2011) 1.2 13 -99.1…”

“…Table 4 shows the results achieved by our classifier (first row) and those of some models available in the literature (Abonyi Table 1 Final parameters of the merged fuzzy model for the classification of solar energy production. 3 4 -96.11 ''HFP-2 n tree'' (Ait Kbir et al, 2000) 392 4 -95.83 HGBML (Ishibuchi et al, 2005a) 10 4 -94.67 FEBFC (Lee et al, 2001) -4 -96.7 HCGA 3.3 --96.22 FH-GBML -4 12 97.33 ''TSK-GA'' (Setnes & Roubos, 2000) 3 4 12 99.4 SoHyFIS-Yager (Tung et al, 2012) 16 del Jesus et al, 2004;Fernández et al, 2009;Guillaume & Charnomordic, 2004;Ishibuchi et al, 2005a;Lee et al, 2001;Sanz et al, 2011;Wang & Lee, 2002;Wu et al, 2011). We achieved a mean classification accuracy of 98.32% with a maximum classification accuracy of 98.44% outperforming the other models (except for (Guillaume & Charnomordic, 2004)).…”

“…The class balancing is the following: 59, 71, and 48. 2 2 3 96.82 ABA (del Jesus et al, 2004) ---95.10 HFRBCS (Fernández et al, 2009) -9 -88.24 HFP (Guillaume & Charnomordic, 2004) 7.8 5 -98.4 HGBML (Ishibuchi et al, 2005a) 20 9 -96.68 FEBFC (Lee et al, 2001) -6 -95.14 HCGA 3.1 --96.09 FH-GBML -9 -95.75 SANFIS (Wang & Lee, 2002) 2 9 18 96.3 FNFN (Wu et al, 2011) 1.6 9 -98.3 Table 5 Classification results on Pima Indians diabetes dataset (best result in bold). Table 6 (first row) shows that the results we achieved using only three features outperform all the considered models Guillaume & Charnomordic, 2004;Ishibuchi & Yamamoto, 2005b;Ishibuchi et al, 2005a;Sanz et al, 2011;Setnes & Roubos, 2000;Wang & Lee, 2002;Wu et al, 2011).…”

Knowledge-based Systems and Interestingness Measures: Analysis with Clinical datasets

“…Computer-aided diagnosis has become an inevitable technique in most hospitals and medical centers. Several frameworks and systems such as expert system [1], medical diagnostic system [2], integrated healthcare systems [3], emergency medical decision support system (MDSS), clinical decision support system [4], clinical recommender system [5], prediction system, and various other machine learning models [6][7][8][9], are used for efficient diagnosis and prognosis. These systems are preferred more than human experts due to factors such as efficient response, consistency, targetspecificity, permanence, and sometimes even multi-domain functionality.…”