Proceedings of the 1st ACM International Health Informatics Symposium
| Hi-index | 0.00 |
Increasing availability of multi-scale physiological data opens new horizons for quantitative modeling in biomedical applications. However, practical limitations of existing approaches include both the low accuracy of the simplified analytical models and empirical expert-defined rules and the insufficient interpretability and stability of the pure data-driven models. Recently it was shown that generic boosting-based frameworks can be successfully used to address these challenges of quantitative modeling in financial applications. Boosting and similar ensemble learning techniques are capable of discovering robust multi-component meta-models from a collection of existing and well-understood base models. Accuracy and stability of such interpretable ensembles of complementary models are often significantly higher than those of the single models. Here we establish the plausibility that this ensemble learning approach can overcome such challenges also in biomedical applications.