Deep Learning-Based Multi-Modal Ensemble Classification Approach for Human Breast Cancer Prognosis

Ensemble models based on deep learning have made significant contributions to the medical field, particularly in area of disease prediction. Breast cancer is a highly aggressive with high mortality rate. Timely and effective prediction breast can reduce risk it progressing later stages need for unnecessary medications. While previous studies focused predicting using single-modal datasets, multi-modal datasets that include gene expression (gene exp), clinical, copy number variation (CNV) data become available recent years predictive model development. However, despite multiple prediction, designed are typically homogeneous neural networks. This article proposes heterogeneous learning-based ensemble data. The consists three phases: feature extraction, stacked set creation, extracted features as input stacked-based random forest algorithm For convolutional networks (CNNs) used clinical data, (DNNs) CNV from CNNs DNNs create comprehensive set. simulation results demonstrate superiority proposed framework terms accuracy compared uni-modal model-multi-modal frameworks.