Construction of a Quasi-Linear Ensemble of Algorithms for Diagnosing Tumors Detected from Mammography Images
Abstract
For breast cancer treatment to be effective, cancers found from mammography images must be accurately and promptly diagnosed. This paper presents the construction of a quasi-linear ensemble of algorithms specifically constructed for this purpose. The suggested ensemble seeks to improve diagnostic robustness and accuracy by utilizing the advantages of several machine-learning models. Quick and accurate diagnoses are ensured by the quasi-linear ensemble's ability to efficiently compute and integrate the various model outputs. Comprehensive testing on openly accessible mammography datasets like CBIS-DDSM and MIAS shows that the suggested ensemble performs better than single algorithms and traditional ensemble techniques. Significant improvements in sensitivity, specificity, and overall diagnostic accuracy are shown by the results.