A Cascade Deep Forest Model for Breast Cancer Subtype Classification Using Multi-Omics Data

Automated diagnosis systems aim to reduce the cost of while maintaining same efficiency. Many methods have been used for breast cancer subtype classification. Some use single data source, others integrate many sources, case that results in reduced computational performance as opposed accuracy. Breast data, especially biological is known its imbalance, with lack extensive amounts histopathological images data. Recent studies shown cascade Deep Forest ensemble model achieves a competitive classification accuracy compared other alternatives, such general learning and conventional deep neural networks (DNNs), imbalanced training sets, through hyper-representations using decision trees. In this work, employed classify subtypes, IntClust Pam50, multi-omics datasets different configurations. The obtained recorded an 83.45% 5 subtypes 77.55% 10 subtypes. significance work it gene expression alone classifier comparable techniques higher performance, where time about s 7