We demonstrate the advantages of using Bayesian multi layer perceptron (MLP) neural networks for image analysis. The Bayesian approach provides consistent way to do inference by c...
Bias/variance analysis is a useful tool for investigating the performance of machine learning algorithms. Conventional analysis decomposes loss into errors due to aspects of the le...
Extensive experimental evidence is required to study the impact of text categorization approaches on real data and to assess the performance within operational scenarios. In this ...
Roberto Basili, Alessandro Moschitti, Maria Teresa...
Several methods to select variables that are subsequently used in discriminant analysis are proposed and analysed. The aim is to find from among a set of m variables a smaller sub...
Background: The amount of information stemming from proteomics experiments involving (multi dimensional) separation techniques, mass spectrometric analysis, and computational anal...