Skip to main content
Figure 1 | BMC Bioinformatics

Figure 1

From: A genetic algorithm-Bayesian network approach for the analysis of metabolomics and spectroscopic data: application to the rapid identification of Bacillus spores and classification of Bacillus species

Figure 1

Work flow for the data analysis. A genetic algorithm is used for feature selection. The GA's fitness function is the predictive accuracy, on the validation data, given by Bayesian network model built for each individual candidate solution. At the end of the GA run, a new BN model is built for the best GA solution using the combined training and validation data. This new BN model is then tested on the previously untouched test data. The classification results described in this paper are the average predictive accuracy over 10 test folds.

Back to article page