Explore the vast range of titles available.

Management and efficient operations in critical infrastructures such as smart grids take huge advantage of accurate power load forecasting, which, due to its non‐linear nature, remains a challenging task. Recently, deep learning has emerged in the machine learning field achieving impressive performance in a vast range of tasks, from image classification to machine translation. Applications of deep learning models to the electric load forecasting problem are gaining interest among researchers as well as the industry, but a comprehensive and sound comparison among different—also traditional—architectures is not yet available in the literature. This work aims at filling the gap by reviewing and experimentally evaluating four real world datasets on the most recent trends in electric load forecasting, by contrasting deep learning architectures on short‐term forecast (one‐day‐ahead prediction). Specifically, the focus is on feedforward and recurrent neural networks, sequence‐to‐sequence models and temporal convolutional neural networks along with architectural variants, which are known in the signal processing community but are novel to the load forecasting one.

Convolutional neural networks (CNNs) are the specific architecture of feed‐forward artificial neural networks. It is the de‐facto standard for various operations in machine learning and computer vision. To transform this performance towards the task of network anomaly detection in cyber‐security, this study proposes a model using one‐dimensional CNN architecture. The authors' approach divides network traffic data into transmission control protocol (TCP), user datagram protocol (UDP), and OTHER protocol categories in the first phase, then each category is treated independently. Before training the model, feature selection is performed using the Chi‐square technique, and then, over‐sampling is conducted using the synthetic minority over‐sampling technique to tackle a class imbalance problem. The authors' method yields the weighted average f‐score 0.85, 0.97, 0.86, and 0.78 for TCP, UDP, OTHER, and ALL categories, respectively. The model is tested on the UNSW‐NB15 dataset.

The Stock Market is one of the most active research areas, and predicting its nature is an epic necessity nowadays. Predicting the Stock Market is quite challenging, and it requires intensive study of the pattern of data. Specific statistical models and artificially intelligent algorithms are needed to meet this challenge and arrive at an appropriate solution. Various machine learning and deep learning algorithms can make a firm prediction with minimised error possibilities. The Artificial Neural Network (ANN) or Deep Feed‐forward Neural Network and the Convolutional Neural Network (CNN) are the two network models that have been used extensively to predict the stock market prices. The models have been used to predict upcoming days' data values from the last few days' data values. This process keeps on repeating recursively as long as the dataset is valid. An endeavour has been taken to optimise this prediction using deep learning, and it has given substantial results. The ANN model achieved an accuracy of 97.66%, whereas the CNN model achieved an accuracy of 98.92%. The CNN model used 2‐D histograms generated out of the quantised dataset within a particular time frame, and prediction is made on that data. This approach has not been implemented earlier for the analysis of such datasets. As a case study, the model has been tested on the recent COVID‐19 pandemic, which caused a sudden downfall of the stock market. The results obtained from this study was decent enough as it produced an accuracy of 91%.

The adiabatic capacitive artificial neuron (ACAN) has been previously shown to offer the potential for ultra‐low power computation in full custom analogue ASIC designs. However, it did not consider how a real‐world, software‐trained, artificial neuron (AN) could be mapped robustly onto the circuit. In this paper, we describe how an AN, with positive‐valued weights, bias and a binary activation function, can be mapped directly and precisely onto an ACAN. The functional equivalence and properties of the mapping are demonstrated with ANs extracted from an artificial neural network (ANN) trained against a binarized MNIST dataset using an open‐source software AI framework.

Activation‐descent regularization is a crucial approach in input optimization for ReLU networks, but traditional methods face challenges. Converting discrete activation patterns into differentiable forms introduces half‐space division, high computational complexity, and instability. We propose a novel local descent regularization method based on a network of arbitrary odd functions, which unifies half‐space processing, simplifies expression, reduces computational complexity, and enriches the expression of the activation descent regularization term. Furthermore, by selecting an arbitrary differentiable odd function, we can derive an exact gradient descent direction, solving the non‐differentiability problem caused by the non‐smooth nature of ReLU, thus improving optimization efficiency and convergence stability. Experiments demonstrate the competitive performance of our approach, particularly in adversarial learning applications. This work contributes to both theory and practice of regularization for input optimization. We propose a novel unified odd‐descent regularization for input optimization, which improves the effectiveness and efficiency of input‐based optimization tasks in deep learning. Our approach exploits odd‐function activation patterns, unifies the treatment of half‐spaces, simplifies the optimization process, and enriches the expression of the activation descent regularization term.

The performance of a stator current-based model reference adaptive systems (MRAS) speed estimator for sensorless induction motor drives is investigated in this study. The measured stator currents are used as a reference model for the MRAS observer to avoid the use of a pure integrator. A two-layer, online-trained neural network stator current observer is used as the adaptive model for the MRAS estimator which requires the rotor flux information. This can be obtained from the voltage or current models, but instability and dc drift can downgrade the overall observer performance. To overcome these problems of rotor flux estimation, an off-line trained multilayer feed-forward neural network is proposed here as a rotor flux observer. Hence, two networks are employed: the first is online trained for stator current estimation and the second is off-line trained for rotor flux estimation. Sensorless operation for the proposed MRAS scheme using current model and neural network rotor flux observers are investigated based on a set of experimental tests in the low-speed region. Using a neural network rotor flux observer to replace the current model is shown to solve the stability problem in the low-speed regenerating mode of operation.

The increasing coupling between the physical and communication layers in the cyber-physical system (CPS) brings up new challenges in system monitoring and control. Smart power grids with the integration of information and communication technologies are one of the most important types of CPS. Proper monitoring and control of the smart grid are highly dependent on the transient stability assessment (TSA). Effective TSA can provide system operators with insightful information on stability statuses and causes under various contingencies and cyber-attacks. In this study, a real-time stability condition predictor based on a feedforward neural network is proposed. The conjugate gradient backpropagation algorithm and Fletcher–Reeves updates are used for training, and the Kohonen learning algorithm is utilised to improve the learning process. By real-time assessment of the network features based on the minimum redundancy maximum relevancy algorithm, the proposed method can successfully predict transient stability and out of step conditions for the network and generators, respectively. Simulation results on the IEEE 39-bus test system indicate the superiority of the proposed method in terms of accuracy, precision, false positive rate, and true positive rate.

Among the recently insulating materials broadly utilized in high voltage outdoor insulation, silicone rubber (SiR) has gotten the foremost consideration. Actually, SiR is becoming an efficient countermeasure to insulator contamination issues. To enhance different properties of polymeric materials, micro- and nanofillers have been used for dielectric applications. In this study, micron-sized titanium dioxide (TiO2) and nano-sized TiO2 fillers were added to the SiR matrix to improve electrical and mechanical properties. Dielectric strength, tensile strength, and elongation at break tests were monitored. Also, a scanning electron microscope was carried out. The samples were prepared by mixing micro-TiO2 into SiR with the content of 0, 10, 20, 30, and 40 wt% and also mixing nano-TiO2 into SiR with the content of 0, 1, 3, 5, and 7 wt%. A feed-forward neural network technique was used to estimate the dielectric strength in different conditions and different percentages of fillers. Adding nano TiO2 filler enhances the electrical and mechanical properties of SiR composites. SiR with 5 wt% nano TiO2 showed the best improvement in electrical and mechanical properties.

In recent years, instance-level-image retrieval has attracted massive attention. Several researchers proposed that the representations learned by convolutional neural network (CNN) can be used for image retrieval task. In this study, the authors propose an effective feature encoder to extract robust information from CNN. It consists of two main steps: the embedding step and the aggregation step. Moreover, they apply the multi-task loss function to train their model in order to make the training process more effective. Finally, this study proposes a novel representation policy that encodes feature vectors extracted from different layers to capture both local patterns and semantic concepts from deep CNN. They call this ‘multi-level-image representation’, which could further improve the performance. The proposed model is helpful to improve the retrieval performance. For the sake of comprehensively evaluating the performance of their approach, they conducted ablation experiments with various convolutional NN architectures. Furthermore, they apply their approach to a concrete challenge – Alibaba large-scale search challenge. The results show that their model is effective and competitive.

Among the recently insulating materials broadly utilized in high voltage outdoor insulation, silicone rubber (SiR) has gotten the foremost consideration. Actually, SiR is becoming an efficient countermeasure to insulator contamination issues. To enhance different properties of polymeric materials, micro- and nanofillers have been used for dielectric applications. In this study, micron-sized titanium dioxide (TiO(2)) and nano-sized TiO(2) fillers were added to the SiR matrix to improve electrical and mechanical properties. Dielectric strength, tensile strength, and elongation at break tests were monitored. Also, a scanning electron microscope was carried out. The samples were prepared by mixing micro-TiO(2) into SiR with the content of 0, 10, 20, 30, and 40 wt% and also mixing nano-TiO(2) into SiR with the content of 0, 1, 3, 5, and 7 wt%. A feed-forward neural network technique was used to estimate the dielectric strength in different conditions and different percentages of fillers. Adding nano TiO(2) filler enhances the electrical and mechanical properties of SiR composites. SiR with 5 wt% nano TiO(2) showed the best improvement in electrical and mechanical properties.