Fig. 1
From: Deep learning-based cancer survival prognosis from RNA-seq data: approaches and evaluations
Neural network architectures of three Deep Learning-based models. a Cox-nnet with a single hidden layer; b DeepSurv with multiple hidden layers having consistent dimensions; c AECOX with multiple hidden layers in the both encoder and decoder part. Last hidden layers in all models were indicated in orange and were connect to a Cox regression neural networks with hazard ratios as the outputs