Embodiments described herein relate to methods and apparatuses for using a Bayesian graph neural network, B-GNN, to determine recommended configurations for a plurality of nodes, wherein the B-GNN comprises the plurality of nodes and a plurality of edges, and wherein the B-GNN is trained to predict the one or more performance metrics associated with the plurality of nodes based on configurations applied to the plurality of nodes. A computer implemented method comprises performing an optimization process utilizing the B-GNN based on a process execution type to determine the recommended configurations for the plurality of nodes.

Graph Neural Networks (GNNs) have rapidly emerged as powerful tools for modeling complex data structures, particularly in the context of telecommunications, chemistry and social networking. Explainability in GNNs holds essential significance as it empowers stakeholders to gain insights into the inner workings of these complex models, fostering trust and transparency in decision-making processes. In this publication, we address the problem of determining how important is each neighbor for the GNN when classifying a node and how to measure the performance for this specific task. To do this, various known explainability methods are reformulated to calculate the neighbor importance and four new metrics, that aid in determining an explainability method’s reliability, are presented. Our results show that there is almost no difference between the explanations provided by gradient-based techniques in the GNN domain, in contrast to other areas of deep learning where this is an active area of research. This means that efforts in this direction may not produce such promising results for GNNs. In addition, many explainability techniques failed to identify important neighbors when GNNs without self-loops are used (The code to replicate our findings will be available here: https://github.com/EricssonResearch/gnn-neighbors-xai).

Saliency maps have become one of the most widely used interpretability techniques for convolutional neural networks (CNN) due to their simplicity and the quality of the insights they provide. However, there are still some doubts about whether these insights are a trustworthy representation of what CNNs use to come up with their predictions. This paper explores how rescuing the sign of the gradients from the saliency map can lead to a deeper understanding of multi-class classification problems. Using both pretrained and trained from scratch CNNs we unveil that considering the sign and the effect not only of the correct class, but also the influence of the other classes, allows to better identify the pixels of the image that the network is really focusing on. Furthermore, how occluding or altering those pixels is expected to affect the outcome also becomes clearer.

This paper presents a novel approach to electricity price forecasting (EPF) using a pure Transformer model. As opposed to other alternatives, no other recurrent network is used in combination to the attention mechanism. Hence, showing that the attention layer is enough for capturing the temporal patterns. The paper also provides fair comparison of the models using the open-source EPF toolbox and provide the code to enhance reproducibility and transparency in EPF research. The results show that the Transformer model outperforms traditional methods, offering a promising solution for reliable and sustainable power system operation.