Memory-Efficient, Accurate and Early Diagnosis of Diabetes Through a Machine Learning Pipeline Employing Crow Search-Based Feature Engineering and a Stacking Ensemble

The early diagnosis of diabetes helps in avoiding the major risks associated with disorder. proposed research involves design a machine learning pipeline which generates most representative feature subset minimal size that predicts onset Diabetes highest accuracy. It employs novel dataset is gender-neutral and enough unlike well-known PID dataset. pipelines involve multiple engineering to generate reduced fed into heterogeneous classifiers. selection as well extraction. former uses ANOVA filter Crow Search Optimization algorithm. latter Singular Value Decomposition. classification performed on preprocessed using wide range classifiers like Naive Bayes’, Logistic Regression, K-Nearest Neighbor, Decision Trees, Support Vector Machine, Random Forest, AdaBoost, GradientBoost base learners followed by their stacking ensemble. performance evaluation each done through Repeated Stratified K-fold Cross Validation metrics accuracy, precision, recall, F1 Score area under Receiver Operating Characteristic curve. For pipeline, number features varies accuracy 98.4% achieved algorithm ensemble A comparative analysis recent related work same shows set outperform improved size.