11 papers
Learning with the Nash-Sutcliffe loss
Hristos Tyralis, Georgia Papacharalampous
The Nash-Sutcliffe efficiency () is a widely used, positively oriented relative measure for evaluating forecasts across multiple time series. However, it lacks a decisi…
Kling-Gupta linear regression
Hristos Tyralis, Georgia Papacharalampous
Kling-Gupta efficiency () is a model performance evaluation metric widely used in hydrology, but its properties as a statistical estimator have remained unexplored. W…
Variable transformations in consistent loss functions
Hristos Tyralis, Georgia Papacharalampous
The empirical use of variable transformations within (strictly) consistent loss functions is widespread, yet a theoretical understanding is lacking. To address this gap, we develop…
Loss functions arising from the index of agreement
Hristos Tyralis, Georgia Papacharalampous
We examine the theoretical properties of the index of agreement loss function , the negatively oriented counterpart of Willmott's index of agreement, a common metric in enviro…
Ensemble learning for uncertainty estimation with application to the correction of satellite precipitation products
Georgia Papacharalampous, Hristos Tyralis, Nikolaos Doulamis +1
Predictions in the form of probability distributions are crucial for effective decision-making. Quantile regression enables such predictions within spatial prediction settings that…
Deep Huber quantile regression networks
Hristos Tyralis, Georgia Papacharalampous, Nilay Dogulu +1
Typical machine learning regression applications aim to report the mean or the median of the predictive probability distribution, via training with a squared or an absolute error s…