2020
DOI: 10.1002/for.2652
|View full text |Cite
|
Sign up to set email alerts
|

Diagnosis of diabetes mellitus using artificial neural network and classification and regression tree optimized with genetic algorithm

Abstract: Diabetes mellitus is one of the most important public health problems affecting millions of people worldwide. An early and accurate diagnosis of diabetes mellitus has critical importance for the medical treatments of patients. In this study, first, artificial neural network (ANN) and classification and regression tree (CART)-based approaches are proposed for the diagnosis of diabetes.Hybrid ANN-GA and CART-GA approaches are then developed using a genetic algorithm (GA) to improve the classification accuracy of… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
11
0
1

Year Published

2021
2021
2024
2024

Publication Types

Select...
5
1

Relationship

1
5

Authors

Journals

citations
Cited by 15 publications
(12 citation statements)
references
References 26 publications
0
11
0
1
Order By: Relevance
“…The Gini index of each attribute can be calculated using Equation 3. GiniA()T=||T1TitalicGini()T1+||T2TitalicGini()T2 The attribute that has the maximum of the degradation in impurity is picked as the splitting attribute (Pekel Özmen & Özcan, 2020). The reduction in impurity is obtained from Equation 4.…”
Section: Methodsmentioning
confidence: 99%
“…The Gini index of each attribute can be calculated using Equation 3. GiniA()T=||T1TitalicGini()T1+||T2TitalicGini()T2 The attribute that has the maximum of the degradation in impurity is picked as the splitting attribute (Pekel Özmen & Özcan, 2020). The reduction in impurity is obtained from Equation 4.…”
Section: Methodsmentioning
confidence: 99%
“…The null values (0) of the Indian PIMA, Glucose (5), BlooPressure(35), SkinThickness(227), Insulin (374) and BMI (11) dataset characteristics are identified as missing values (Table 3). But this study was conducted with data containing missing values in their original state, considered as null values.…”
Section: Presentation Of the Datasetmentioning
confidence: 99%
“…This may explain the fact that these algorithms outperform others in comparative studies. The results of better estimators that are DT in [11]RF in [12], Adaboost in [5], LightGBM in [16] and XGBoost in [13]give us comfort in the choice of tree methods to conduct our study. As the aim of the study is to detect almost all diabetics and minimize the false negative rate, we prefer the classifier that achieves good sensitivity.…”
Section: K-fold (K=12)mentioning
confidence: 99%
See 1 more Smart Citation
“…Neural networks are widely applied in various types of classification and regression tasks (Borovkova & Tsiamas, 2019; Bucci, 2020; Kraft et al, 2020; Pekel Ozmen & Ozcan, 2020) because of their high learning capability and strong approximation ability for nonlinear series. However, iterative learning methods, such as backpropagation (BP) neural network, radial basis functional (RBF) network, long short‐term memory (LSTM), and others, have the shortcomings of long training time (Tang et al, 2018), easily falling into local optima, and slow convergence speeds (Chitsazan et al, 2019).…”
Section: Introductionmentioning
confidence: 99%