Genetic algorithm wrapped Bayesian network feature selection applied to differential diagnosis of erythemato-squamous diseases

A mobile/desktop medical application for automatic differential diagnosis of psoriasis lesions

“…The classification of erythemato-squamous diseases in one of the six types of diseases has been intensely studied [1][2][3][4][5][6][7][8][9][10][11][12]. Nevertheless, in these studies partial or all histopathological features were used for the classification purpose.…”

“…Having a large number of features can lead to "over-fitting" [11]. This fact has led to the elimination of some features from the Dermatology dataset by researchers that studied it, in order to increase the classifier's performance and accuracy [8], [12].…”

“…The model achieved 98.91% classification accuracy on the Dermatology dataset. The most recent algorithm applied on this dataset is the genetic wrapped Bayesian network feature selection algorithm proposed in [12], for which 99.2% classification accuracy was achieved. The overall results are summarized in Table I.…”

“…Polat [6] kNN 97.57 Polat [6] Decision tree 99 Polat [7] C4.5 96.71 Karabatak [8] AR-NN 98.61 Lekkas [9] Fuzzy classification 97.55 Xie [10] SVM 98.61 Abdi [11] PSO-SVM 98.91 Ozcift [12] Genetic wrapped Bayesian network…”

Supervised Feature Selection for Diagnosis of Coronary Artery Disease Based on Genetic Algorithm

“…The algorithm was compared also with two wrapper FS methods (BFS and SFFS). The classification accuracies demonstrate the effectiveness of the features subset selected by GA. Setiawan et al [10] Debabrata et al [37] Özçift et al [38] Figure 3. Analysis of different Systems…”