Predicting good probabilities with supervised learning
ICML '05 Proceedings of the 22nd international conference on Machine learning
Pareto evolutionary neural networks
IEEE Transactions on Neural Networks
Conditional Density Estimation with Class Probability Estimators
ACML '09 Proceedings of the 1st Asian Conference on Machine Learning: Advances in Machine Learning
Making good probability estimates for regression
ECML'06 Proceedings of the 17th European conference on Machine Learning
Hi-index | 0.00 |
The two objectives when estimating the conditional density function in a regression problem are to maximise sharpness (the density rewarded to the actual observation), while maintaining calibration (the empirical validity of the probability estimates). In this paper we outline a process of optimisation that maximises both these objectives simultaneously to make better probabilistic predictions.