Machine learning approach towards explaining water quality dynamics in an urbanised river

Utilizing machine learning, we find that boosted trees outperform GAM and accurately describe water quality dynamics with less than 1% error.