SSH-Net: A Deep Neural Network for Predicting Failure Time Distribution Functions under Competing Risks with Application to GPU Data
Researchers have developed SSH-Net, a novel deep neural network designed to predict failure time distributions in systems with competing risks, such as GPUs. This Structured Segmented Hazard Deep Neural Network associates network structure with data structure, allowing different data groups to influence predictions through separate sub-networks. The model outputs cause-specific hazard functions and has been validated through simulations and applied to Titan GPU failure time data. AI
IMPACT This model could improve reliability and maintenance predictions for complex engineered systems like GPUs.