Advancements in breast cancer detection: Harnessing the power of machine learning for diagnosis
Ishdeep Singla, Sandeep Singh Kang, Divyanshu Kashyap, Paramjot Singh, Ishaan Gupta, Gagan Chaudhary · Computational Methods in Science and Technology · 2024
Breast cancer remains a significant health challenge globally, emphasizing the critical need for improved detection methodologies. The paper&s;s research explores the integration of machine learning techniques for enhancing breast cancer diagnosis accuracy. Leveraging algorithms including random forest, XGBoost, K-nearest neighbour and logistic regression various predictive models are established and evaluated. The primary focus lies on recall as the pivotal evaluation metric, with accuracy, F1-score and precision also considered. To mitigate dimensional inconsistencies, data standardization is employed, followed by feature selection via statistical tests, resulting in an optimal subset of features for model input. The K-nearest neighbour model employs cross-validation to determine the optimal parameters, addressing the challenge of sample imbalance through stratified sampling for training and testing set extraction. The findings not only validate the efficacy of machine learning in breast cancer detection but also underscore the significance of methodological considerations in optimizing predictive models for diagnosis. This research contributes to the ongoing efforts in advancing initial discovery strategies, ultimately improving patient consequences.