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

arXiv: https://arxiv.org/abs/2605.30119 · HTML: https://mlsanity.com/paper/2605.30119/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 27 references that are in the map.
- First posted on arXiv 28 May 2026.

## Builds on (most cited first, 10 of 10)

- [KAN: Kolmogorov-Arnold Networks](https://mlsanity.com/paper/2404.19756/) (2024)
- [ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R](https://mlsanity.com/paper/1508.04409/) (2015)
- [CoxKAN: Kolmogorov-Arnold Networks for Interpretable, High-Performance Survival Analysis](https://mlsanity.com/paper/2409.04290/) (2024)
- [On Explaining Machine Learning Models by Evolving Crucial and Compact Features](https://mlsanity.com/paper/1907.02260/) (2019)
- [Optimal Survival Trees: A Dynamic Programming Approach](https://mlsanity.com/paper/2401.04489/) (2024)
- [Optimal Sparse Survival Trees](https://mlsanity.com/paper/2401.15330/) (2024)
- [Improving the efficiency of GP-GOMEA for higher-arity operators](https://mlsanity.com/paper/2402.09854/) (2024)
- [PISA: An AI Pipeline for Interpretable-by-design Survival Analysis Providing Multiple Complexity-Accuracy Trade-off Models](https://mlsanity.com/paper/2509.22673/) (2025)
- [Multi-modal multi-objective model-based genetic programming to find multiple diverse high-quality models](https://mlsanity.com/paper/2203.13347/) (2022)
- [Thinking Outside the Template with Modular GP-GOMEA](https://mlsanity.com/paper/2505.01262/) (2025)

## Abstract (arXiv)

Survival analysis concerns the task of predicting the time until an event occurs. Often used in the medical field, survival analysis deals with incomplete (i.e., censored) data, for instance, from patients who did not experience the event during the duration of the study. For practical use, both accuracy and interpretability are important. Survival trees are easy-to-follow survival models that split the patient cohort recursively into discrete patient groups. Whilst survival trees can capture complex relationships, they typically need to grow large, threatening interpretability. Moreover, survival trees are often built using greedy approaches that may overlook globally optimal split combinations, limiting predictive performance. Shallow survival trees require expressive, higher-order feature combinations to achieve competitive accuracy. We therefore use genetic programming to multi-objectively evolve inherently inspectable feature sets and study how they interact with different tree induction strategies. We further introduce an evolutionary approach that jointly optimises the survival tree structure and the non-linear split logic. Our findings demonstrate that evolutionary feature construction improves predictive performance across different tree induction strategies on two real-world datasets and two different survival tree depths. Given its speed and flexible presentation, the multi-objective evolution of entire trees likely holds the most future promise.