Efficient heart disease prediction-based on optimal feature selection using DFCSS and classification by improved...

Area of Science:

• Clinical data analysis
• Machine learning for healthcare
• Cardiovascular disease prediction

Background:

• Cardiovascular disease (CVD) prediction is a significant challenge in clinical data analysis.
• Accurate CVD prediction requires efficient feature selection and robust classification models.

Purpose of the Study:

• To develop an efficient heart disease prediction model using optimal feature selection.
• To enhance the accuracy of cardiovascular disease classification through an integrated approach.

Main Methods:

• Data pre-processing including cleaning, transformation, imputation, and normalization.
• Feature selection using the decision function-based chaotic salp swarm (DFCSS) algorithm.
• Classification using an improved Elman neural network (IENN) optimized by the sailfish optimisation (SFO) algorithm, forming the DFCSS-IESFO model.

Main Results:

• The DFCSS-IESFO model achieved high classification accuracy: 98.7% on the CVD dataset and 98% on the UCI Cleveland heart disease dataset.
• The proposed method demonstrated superior performance compared to support vector machine, K-nearest neighbour, Elman neural network, Gaussian Naive Bayes, logistic regression, random forest, and decision tree classifiers.

Conclusions:

• The DFCSS-IESFO approach is effective for accurate cardiovascular disease prediction.
• Optimal feature selection and optimized neural networks significantly improve heart disease classification accuracy.