Enhancing Breast Cancer Prediction with an Advanced K-Nearest Neighbors (KNN) Algorithm Integrated with Feedback Support Mechanism
Christian Arthur, Kristoko Dwi Hartomo · 2023
Cancer poses a formidable global health challenge characterized by a persistent rise in mortality rates, with breast cancer emerging as the predominant form in 2020. Effectively addressing this heightened mortality necessitates urgent measures for early detection and timely intervention. Within the scientific landscape, numerous antecedent studies have leveraged machine learning methodologies, notably K-Nearest Neighbors (KNN) with feedback support, showcasing a substantial impact on model performance. Particularly noteworthy is the observation that for K-values exceeding 10, models incorporating feedback support demonstrated superior accuracy compared to traditional KNN approaches. Importantly, our Proposed Improved KNN introduced a noteworthy advancement, enhancing accuracy by 2.8%, from 92.3% to 95.1%. This improvement signifies a promising development in the realm of early cancer detection, highlighting the potential efficacy of our proposed approach.