Evolving Features vs Evolving Entire Trees with GP for Interpretable Survival Analysis

2026-05-28 · Source: Machine Learning · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Health & Medical Research · Depth: Advanced, quick

Summary

A new approach utilizes genetic programming (GP) to enhance interpretable survival analysis, addressing limitations of traditional survival trees. While conventional survival trees offer interpretability, they often require significant depth for accuracy, which compromises their clarity, and their greedy construction methods can miss globally optimal splits. This research proposes using GP to multi-objectively evolve inherently inspectable feature sets, examining their interaction with various tree induction strategies. Furthermore, it introduces an evolutionary method that jointly optimizes both the survival tree structure and its non-linear split logic. Findings demonstrate that evolutionary feature construction consistently improves predictive performance across different tree induction strategies on two real-world datasets and two distinct survival tree depths. The full joint evolution method shows the highest potential for generating multiple high-performing, inherently inspectable shallow survival trees.

Key takeaway

For Machine Learning Engineers developing interpretable survival models, you should integrate genetic programming to overcome limitations of traditional greedy tree induction. This approach allows you to evolve expressive feature sets and jointly optimize tree structures, leading to more accurate and inherently inspectable shallow models. Consider applying these evolutionary methods to improve predictive performance on your real-world datasets, especially when balancing model clarity with high accuracy is critical.

Key insights

Genetic programming can jointly optimize survival tree features and structure for improved interpretability and predictive performance.

Principles

• Shallow survival trees need expressive, higher-order features.
• Greedy tree induction may miss globally optimal splits.
• Multi-objective evolution can balance accuracy and interpretability.

Method

Genetic programming multi-objectively evolves inspectable feature sets. An evolutionary approach jointly optimizes survival tree structure and non-linear split logic for improved performance.

In practice

• Apply GP for feature engineering in survival models.
• Consider joint evolution for tree structure and splits.
• Evaluate evolutionary methods on real-world datasets.

Topics

• Survival Analysis
• Genetic Programming
• Interpretable AI
• Feature Engineering
• Evolutionary Computing
• Decision Trees

Best for: AI Scientist, Research Scientist, Machine Learning Engineer

Related on AIssential

• Learning and Naming Subgroups with Exceptional Survival Characteristics — Sysurv identifies exceptional survival subgroups using differentiable, non-parametric methods and interpretable rules.
• L-System Genetic Encoding for Scalable Neural Network Evolution: A Comparison with Direct Matrix Encoding — L-System genetic encoding significantly outperforms direct matrix encoding for neural network evolution in performance and generalization.
• Isomorphic Functionalities between Ant Colony and Ensemble Learning: Part II-On the Strength of Weak Learnability and the Boosting Paradigm — Boosting algorithms and ant colony adaptive recruitment are mathematically isomorphic, both achieving bias reduction through adaptive weighting.
• Every Supervised Machine Learning Model Is Linear — Feature engineering via geometric algebra can enable linear models to outperform complex ensembles like XGBoost.
• Guiding Multi-Objective Genetic Programming with Description Length Improves Symbolic Regression Solutions — Description Length (DL) and Fractional Bayes Factor (FBF) effectively prevent overfitting and bloat in symbolic regression when used with multi-objective genetic programming.
• Distribution-Free Pretraining of Classification Losses via Evolutionary Dynamics — EDL pretrains transferable classification losses using synthetic data and evolutionary optimization with chaotic mutation.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine Learning.