In the last decades, the long-term structural health monitoring of civil structures has been mainly performed using two approaches: model based and data based. The former approach tries to identify damage by relating the monitoring data to the prediction of numerical (e.g., finite-element) models of the structure. The latter approach is data driven, where measured data from a given state condition are compared to the baseline or reference condition. A challenge in both approaches is to make the distinction between the changes of the structural response caused by damage and environmental or operational variability. This issue was tackled here using a hybrid technique that integrates model- and data-based approaches into structural health monitoring. Data recorded in situ under normal conditions were combined with data obtained from finite-element simulations of more extreme environmental and operational scenarios and input into the training process of machine-learning algorithms for damage detection. The addition of simulated data enabled a sharper classification of damage by avoiding false positives induced by wide environmental and operational variability. The procedure was applied to the Z-24 Bridge, for which 1 year of continuous monitoring data were available.

This paper aims to compare the convolutional neural networks (CNNs): ResNet50, InceptionV3, and InceptionResNetV2 tested with and without pre-trained weights on the ImageNet database in order to solve the scene recognition problem. The results showed that the pre-trained ResNet50 achieved the best performance with an average accuracy of 99.82% in training and 85.53% in the test, while the worst result was attributed to the ResNet50 without pre-training, with 88.76% and 71.66% of average accuracy in training and testing, respectively. The main contribution of this work is the direct comparison between the CNNs widely applied in the literature, that is, to enable a better selection of the algorithms in the various scene recognition applications.

The rapid growth of the Internet and technological advances are forcing mobile operators to increasingly invest in network infrastructures. C-RAN and SDN are regarded as enabling technologies that can overcome the limitations faced by operators, by reducing costs, increasing scalability, and paving the way for the next generation of 5G cellular networks. In this paper, an architectural solution based on SDN and computational intelligence is proposed for C-RAN, which can adjust BBU-RRH mapping through network load balancing rules by predicting subjective and objective QoE metrics for UHD video streaming. The simulation results achieved gains between 59% and 129%, in scenarios without activating a new BBU and scenarios that involve activating a new BBU, respectively.

Pneumonia is one of the most common medical problems in clinical practice and is the leading fatal infectious disease worldwide. According to the World Health Organization, pneumonia kills about 2 million children under the age of 5 and is constantly estimated to be the leading cause of infant mortality, killing more children than AIDS, malaria, and measles combined. A key element in the diagnosis is radiographic data, as chest x-rays are routinely obtained as a standard of care and can aid to differentiate the types of pneumonia. However, a rapid radiological interpretation of images is not always available, particularly in places with few resources, where childhood pneumonia has the highest incidence and mortality rates. As an alternative, the application of deep learning techniques for the classification of medical images has grown considerably in recent years. This study presents five implementations of convolutional neural networks (CNNs): ResNet50, VGG-16, InceptionV3, InceptionResNetV2, and ResNeXt50. To support the diagnosis of the disease, these CNNs were applied to solve the classification problem of medical radiographs from people with pneumonia. InceptionResNetV2 obtained the best recall and precision results for the Normal and Pneumonia classes, 93.95% and 97.52% respectively. ResNeXt50 achieved the best precision and f1-score results for the Normal class (94.62% and 94.25% respectively) and the recall and f1-score results for the Pneumonia class (97.80% and 97.65%, respectively).