Noisy data facilitates researcher’s investigation of breast cancer gene expression

Researchers report on the use of denoising autoencoders (DAs) to effectively extract key biological principles from gene expression data and summarize them into constructed features with convenient properties.