Training Data Selection for Machine Learning-Enhanced Monte Carlo Simulations in Structural Dynamics

The evaluation of structural response constitutes a fundamental task in the design of ground-excited structures. In this context, the Monte Carlo simulation is a powerful tool to estimate the response statistics of nonlinear systems, which cannot be represented analytically. Unfortunately, the number of samples which is required for estimations with high confidence increases disproportionally to obtain a reliable estimation of low-probability events. As a consequence, the Monte Carlo simulation becomes a non-realizable task from a computational perspective. We show that the application of machine learning algorithms significantly lowers the computational burden of the Monte Carlo method. We use artificial neural networks to predict structural response behavior using supervised learning. However, one shortcoming of supervised learning is the inability of a sufficiently accurate prediction when extrapolating to data the neural network has not seen yet. In this paper, neural networks predict the response of structures subjected to non-stationary ground excitations. In doing so, we propose a novel selection process for the training data to provide the required samples to reliably predict rare events. We, finally, prove that the new strategy results in a significant improvement of the prediction of the response statistics in the tail end of the distribution.

An Empirical Study of Training Data Selection Methods for Ranking-Oriented Cross-Project Defect Prediction

Ranking-oriented cross-project defect prediction (ROCPDP), which ranks software modules of a new target industrial project based on the predicted defect number or density, has been suggested in the literature. A major concern of ROCPDP is the distribution difference between the source project (aka. within-project) data and target project (aka. cross-project) data, which evidently degrades prediction performance. To investigate the impacts of training data selection methods on the performances of ROCPDP models, we examined the practical effects of nine training data selection methods, including a global filter, which does not filter out any cross-project data. Additionally, the prediction performances of ROCPDP models trained on the filtered cross-project data using the training data selection methods were compared with those of ranking-oriented within-project defect prediction (ROWPDP) models trained on sufficient and limited within-project data. Eleven available defect datasets from the industrial projects were considered and evaluated using two ranking performance measures, i.e., FPA and Norm(Popt). The results showed no statistically significant differences among these nine training data selection methods in terms of FPA and Norm(Popt). The performances of ROCPDP models trained on filtered cross-project data were not comparable with those of ROWPDP models trained on sufficient historical within-project data. However, ROCPDP models trained on filtered cross-project data achieved better performance values than ROWPDP models trained on limited historical within-project data. Therefore, we recommended that software quality teams exploit other project datasets to perform ROCPDP when there is no or limited within-project data.