Confidence estimation methods for neural networks : a practical comparison

Feed-forward neural networks (Multi-Layered Perceptrons) are used widely in real-world regression or classification tasks. A reliable and practical measure of prediction "confidence" is essential in real-world tasks. This paper compares three approachesto prediction confidenceestimation, using both artificial and real data. The three methods are maximum likelihood, approximate Bayesian and bootstrap. Both noiseinherent to the dataand model uncertaintyare considered.

• Type: Conference
• Year: 2000
• Where: ESANN
• Authors: Georgios Papadopoulos, Peter J. Edwards, Alan F. Murray
• Keywords: ESANN 2000, ESANN 2007, Feed-forward Neural Networks, Real-world Regression, Real-world Tasks