Symbolic regression outperforms other models for small data sets
This addresses the challenge of limited data availability in health science and other domains, offering an incremental improvement in model performance for small datasets.
The study tackled the problem of overfitting in traditional machine learning models like random forests and gradient boosting on small datasets of 250 observations, showing that symbolic regression generalizes better with a higher R2 in 132 out of 240 cases and performs best in 184 out of 240 cases when compared to interpretable models.
Machine learning is often applied in health science to obtain predictions and new understandings of complex phenomena and relationships, but an availability of sufficient data for model training is a widespread problem. Traditional machine learning techniques, such as random forests and gradient boosting, tend to overfit when working with data sets of only a few hundred observations. This study demonstrates that for small training sets of 250 observations, symbolic regression generalises better to out-of-sample data than traditional machine learning frameworks, as measured by the coefficient of determination R2 on the validation set. In 132 out of 240 cases, symbolic regression achieves a higher R2 than any of the other models on the out-of-sample data. Furthermore, symbolic regression also preserves the interpretability of linear models and decision trees, an added benefit to its superior generalisation. The second best algorithm was found to be a random forest, which performs best in 37 of the 240 cases. When restricting the comparison to interpretable models, symbolic regression performs best in 184 out of 240 cases.