Rock-Burst Occurrence Prediction Based on Optimized Naïve Bayes Models

Rock-burst is a common failure in hard rock related projects civil and mining construction therefore, proper classification prediction of this phenomenon interest. This research presents the development optimized naïve Bayes models, predicting rock-burst failures underground projects. The models were using four weight optimization techniques including forward, backward, particle swarm optimization, evolutionary. An evolutionary random forest model was developed to identify most significant input parameters. maximum tangential stress, elastic energy index, uniaxial tensile stress then selected by feature selection technique (i.e., forest) develop models. performance assessed various criteria as well simple ranking system. results showed that effective improving accuracy for (cumulative = 21), while backward worst 11). All identified parameter failures. demonstrate may improve algorithms occurrence.