MENSA: A Multi-Event Network for Survival Analysis with Trajectory-based Likelihood Estimation

2024-09-10

AI Summary

Existing multi-event survival analysis methods typically output risk scores, limiting direct event-time prediction and neglecting temporal dependencies among events. To address this, we propose a trajectory-based likelihood estimation framework that jointly models the occurrence timing and feature representations of non-exclusive, statistically dependent events. Our approach employs deep neural networks for end-to-end time regression, explicitly encoding event ordering via sequential modeling, and replaces the conventional risk-score paradigm with direct minimization of negative log-likelihood over predicted event times. The framework unifies single-event, competing-risk, and multi-event settings within a single architecture. On clinical benchmarks, it achieves significant improvements in temporal prediction accuracy (MAE reduced by 18.3%) and discriminative performance (C-index increased by 0.042), while requiring fewer parameters and lower computational overhead than state-of-the-art baselines.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Reasoning under Uncertainty: Relational Probabilistic Models
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

A multi-event survival model predicts the time until an instance experiences each of several different events, given the instance's description. Unlike competing-event models, the events here are not mutually exclusive and often exhibit statistical dependencies. Existing approaches for multi-event survival analysis have generally been limited, most focusing on producing simple risk scores for each event, rather than the time-to-event itself. To overcome these issues, we introduce MENSA, a novel deep learning approach for multi-event survival analysis. MENSA jointly learns representations of the input features while capturing the complex dependence structure among events. In practice, it attempts to optimize the sum of the traditional negative log-likelihood across events and a novel trajectory-based likelihood, which encourages the model to learn the temporal order in which events occur. Experiments on real-world clinical datasets show that MENSA consistently gives good discrimination performances and accurate time-to-event predictions in single-event, competing-risk, and multi-event problems. Additionally, MENSA is more computationally-efficient, requiring fewer parameters and FLOPs than multiple state-of-the-art survival baselines when applied to large-dimensional datasets (more than 100 features).

Problem

• Predicts time-to-multiple-events
• Captures event dependencies
• Improves computational efficiency

Innovation

• Deep learning for multi-event survival
• Trajectory-based likelihood estimation
• Efficient parameter optimization