Fig. 2From: Deep learning-based cancer survival prognosis from RNA-seq data: approaches and evaluationsa, b: Performance comparisons between three Deep Learning-based models across 12 TCGA (The Cancer Genome Atlas) cancers. a concordance index; b p-value of log-rank test (in −log10 scale). c, d: Performance comparisons between three Deep Learning-based models and three traditional machine learning models across 12 TCGA (The Cancer Genome Atlas) cancers. c concordance index; d p-value of log-rank test (in −log10 scale). Cancers were sorted based on averaged concordance index across models and experiments. For detailed cancer names, please refer to the Additional file 1Back to article page