bioMoR applies a biology-guided Mixture-of-Recursions framework to gene and pathway-level genomic learning, improving av...

bioMoR applies a biology-guided Mixture-of-Recursions framework to gene and pathway-level genomic learning, improving average macro-F1 by 8.2 percentage points and balanced accuracy by 7.1 percentage points over baselines while using 75 percent fewer parameters and up to 58 percent fewer FLOPsSource: arXiv cs.AIhttps://arxiv.org/abs/2608.06727#MachineLearning