Finally, the comparison of CRNet with UBS against proposed CRNet demonstrates highly significant improvements in accuracy, precision, F1-score, and ROC-AUC with F-values of 44.882, 171.204, 236.391, and 44.596 respectively, all with p-values of 0.00000, while recall shows an insignificant difference with an F-value of 0.101 and a p-value of 0.75415.
← all excerpts
A 3-tier information fusioned framework featuring explainable deep active optimized CRNet for accurate heart disease prediction.
1
—
—