Robust Framework to Combine Diverse Classifiers Assigning Distributed Confidence to Individual Classifiers at Class Level
Table 1
Overview of datasets used for experimental evaluation.
Dataset
Description
Number of samples
Number of features
Number of classes
IRIS
A flower dataset [59] used as a test bed for comparison of classification techniques.
150
4
3
DiaretDB
A lesion image dataset [51] used for diabetic retinopathy detection methods. These images are processed by [60, 61] to extract feature vector representation of lesions.
3080
15
4
Satimage
A publicly available Satellite Images dataset [59] generated from Landsat scanner image data.
6435
36
6
Heart
A good dataset [59] to test the ML algorithms. In this dataset, each patient is classified as normal and abnormal.
267
23
2
Pima Indian diabetes
It is provided by National Institute of Diabetes and Digestive and Kidney diseases [59]. It is a dataset of all female patients, at least 21 years old.