Browsing School of Computer Science and Statistics by Author "Carney, Michael"
Now showing items 1-5 of 5
-
The Benefits of Using a Complete Probability Distribution when Decision Making: An Example in Anticoagulant Drug Therapy
Carney, Michael; Cunningham, Padraig (Trinity College Dublin, Department of Computer Science, 2005-08-06)In this paper we aim to show how probabilistic prediction of a continuous variable could be more beneficial to a medical practitioner than classification or numeric/point prediction of the same variable in many scenarios. ... -
Calibrating Probability Density Forecasts with Multi-objective Search
Carney, Michael; Cunningham, Padraig (Trinity College Dublin, Department of Computer Science, 2006-02-10)In this paper, we show that the optimization of density forecasting models for regression in machine learning can be formulated as a multi-objective problem.We describe the two objectives of sharpness and calibration and ... -
Evaluating Density Forecasting Models
Carney, Michael; Cunningham, Padraig (Trinity College Dublin, Department of Computer Science, 2006-05-02)Density forecasting in regression is gaining popularity as real world applications demand an estimate of the level of uncertainty in predictions. In this paper we describe the two goals of density forecasting1 sharpness ... -
Improved optimisation of density forecasting models using multi-objective search with applications in risk management
Carney, Michael (Trinity College (Dublin, Ireland). School of Computer Science & Statistics, 2007)Density forecasting is becoming an increasingly popular method of prediction. Density forcasting models produce a probability density function estimate of a future event rather than a standard point estimate. Given that ... -
Predicting Probability Distributions for Surf Height Using an Ensemble of Mixture Density Networks
Carney, Michael; Cunningham, Padraig; Dowling, Jim (Trinity College Dublin, Department of Computer Science, 2006-02-10)There is a range of potential applications of Machine Learning where it would be more useful to predict the probability distribution for a variable rather than simply the most likely value for that variable. In meteorology ...