Publications by authors named "Linda M See"
Article Synopsis
- This paper introduces a symbiotic adaptive neuro-evolutionary algorithm designed to create optimized neural network models specifically for hydrological modeling in the River Ouse catchment.
- It enhances traditional methods by evolving individual neurons and allows experimentation with various objective functions, not just the standard sum squared error, which recent studies indicate may not yield the best models.
- Results demonstrate that these evolved neural networks perform better than conventional models, particularly at longer lead times (6 and 24 hours), and highlight the limitations of using sum squared error as an objective function.
