WTNN: Weibull-Tailored Neural Networks for survival analysis

2025-12-09

AI Summary

Survival prediction for complex systems—such as military vehicles—under dynamic environments is challenging due to reliance on noisy proxy indicators and prevalent right-censored data. Method: This paper proposes a covariate-driven deep Weibull survival model. It explicitly embeds the prior parameter structure of the Weibull distribution into a neural network architecture, enabling differentiable, time-dependent modeling of both scale and shape parameters conditioned on covariates. Crucially, it supports qualitative incorporation of domain knowledge—e.g., known directional effects of critical covariates—enhancing interpretability without sacrificing flexibility. Contribution/Results: Evaluated on real-world military vehicle datasets, the model significantly outperforms state-of-the-art regression-based and black-box deep survival methods. It demonstrates robustness and high predictive accuracy under stringent conditions: high noise in proxy indicators, substantial censoring rates, and strong environmental perturbations. The integration of parametric survival theory with deep learning thus achieves both improved performance and enhanced transparency.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Relational Probabilistic Models
• Intelligent Robots: State Estimation

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

The Weibull distribution is a commonly adopted choice for modeling the survival of systems subject to maintenance over time. When only proxy indicators and censored observations are available, it becomes necessary to express the distribution's parameters as functions of time-dependent covariates. Deep neural networks provide the flexibility needed to learn complex relationships between these covariates and operational lifetime, thereby extending the capabilities of traditional regression-based models. Motivated by the analysis of a fleet of military vehicles operating in highly variable and demanding environments, as well as by the limitations observed in existing methodologies, this paper introduces WTNN, a new neural network-based modeling framework specifically designed for Weibull survival studies. The proposed architecture is specifically designed to incorporate qualitative prior knowledge regarding the most influential covariates, in a manner consistent with the shape and structure of the Weibull distribution. Through numerical experiments, we show that this approach can be reliably trained on proxy and right-censored data, and is capable of producing robust and interpretable survival predictions that can improve existing approaches.

Problem

• Model survival with proxy and censored data
• Extend traditional regression using neural networks
• Incorporate prior knowledge into Weibull survival analysis

Innovation

• Neural network models Weibull distribution parameters
• Incorporates prior knowledge on influential covariates
• Trains on proxy and censored survival data