A Hybrid Machine Learning Approach for Enhanced Prediction of Breast Cancer with Lasso Method for Feature Extraction

Breast cancer diagnosis remains a crucial area of research due to its frequency and impact on global health. Although machine learning approaches have advanced, new approaches are still required to effectively address issues with feature selection, prediction accuracy, and model interpretability. The innovative hybrid machine learning approach shown in this research combines the benefits of Random Forests (RF) and Artificial Neural Networks (ANN) with the ingenious use of the Least Absolute Shrinkage Selection Operator (LASSO) method for feature extraction. This approach attempts to address limitations identified in earlier studies by reducing feature dimensionality, improving prediction accuracy, and enhancing model interpretability in breast cancer diagnosis. The hybrid model achieved an amazing accuracy rate of 98.2% using the chosen attributes and the Random Forest (RF) and Artificial Neural Network (ANN) algorithms.