Power system Frequency Prediction after disturbance based on Deep Learning

Differential evolution-based optimized hierarchical extreme learning machines for fault section diagnosis of large-scale power systems

A non-linear rainfall-runoff model using radial basis function network

Fully connected multi layer neural networks such as Deep Boltzmann Machines (DBM) performs better than fully connected single layer neural networks in image classification tasks and has a smaller number of hidden layer neurons than Extreme Learning Machine (ELM) based fully connected multi layer neural networks such as Multi Layer ELM (MLELM) and Hierarchical ELM (H-ELM) However, ML-ELM and H-ELM has a smaller training time than DBM. This paper introduces a fully connected multi layer neural network referred to as Multi Layer Multi Objective Extreme Learning Machine (MLMO-ELM) which uses a multi objective formulation to pass the label and non-linear information in order to learn a network model which has a similar number of hidden layer parameters as DBM and smaller training time than DBM. The experimental results show that MLMO-ELM outperforms DBM, ML-ELM and H-ELM on OCR and NORB datasets.

Intelligent algorithms using deep learning can help learn feature data with nonlinearity and uncertainty, such as time-series particle concentration data. This paper proposes an improved particle swarm optimization (IPSO) algorithm using nonlinear decreasing weights to optimize the hyperparameters, such as the number of hidden layer neurons, learning rate, and maximum number of iterations of the long short-term memory (LSTM) neural network, to predict the time series for air particulate concentration and capture its data dependence. The IPSO algorithm uses nonlinear decreasing weights to make the inertia weights nonlinearly decreasing during the iteration process to improve the convergence speed and capability of finding the global optimization of the PSO. This study addresses the limitations of the traditional method and exhibits accurate predictions. The results of the improved algorithm reveal that the root means square, mean absolute percentage error, and mean absolute error of the IPSO-LSTM model predicted changes in six particle concentrations, which decreased by 1.59% to 5.35%, 0.25% to 3.82%, 7.82% to 13.65%, 0.7% to 3.62%, 0.01% to 3.55%, and 1.06% to 17.21%, respectively, compared with the LSTM and PSO-LSTM models. The IPSO-LSTM prediction model has higher accuracy than the other models, and its accurate prediction model is suitable for regional air quality management and effective control of the adverse effects of air pollution.

This paper aims to develop an artificial neural network (ANN) to predict the energy consumption and cost of cross laminated timber (CLT) office buildings in severe cold regions during the early stage of architectural design. Eleven variables were selected as input variables including building form and construction variables, and the values of input variables were determined by local building standards and surveys. ANNs were trained by the simulation data and Latin hypercube sampling (LHS) method was used to select training datasets for the ANN training. The best ANN was obtained by analyzing the output variables and the number of hidden layer neurons. The results showed that the ANN with multiple outputs presented better prediction performance than the ANN with single output. Moreover, the number of hidden layer neurons in ANN should be greater than five and preferably 10, and the best mean square error (MSE) value was 1.957 × 103. In addition, it was found that the time of predicting building energy consumption and cost by ANN was 80% shorter than that of traditional building energy consumption simulation and cost calculation method.

Cervical cancer is one of the most common types of cancer among women, which has higher death-rate than many other cancer types. The most common way to diagnose cervical cancer is to analyze images of cervical cells, which is performed using Pap smear imaging test. Early and accurate diagnosis can save the lives of many patients and increase the chance of success of treatment methods. Until now, various methods have been proposed to diagnose cervical cancer based on the analysis of Pap smear images. Most of the existing methods can be divided into two groups of methods based on deep learning techniques or machine learning algorithms. In this study, a combination method is presented, whose overall structure is based on a machine learning strategy, where the feature extraction stage is completely separate from the classification stage. However, in the feature extraction stage, deep networks are used. In this paper, a multi-layer perceptron (MLP) neural network fed with deep features is presented. The number of hidden layer neurons is tuned based on four innovative ideas. Additionally, ResNet-34, ResNet-50 and VGG-19 deep networks have been used to feed MLP. In the presented method, the layers related to the classification phase are removed in these two CNN networks, and the outputs feed the MLP after passing through a flatten layer. In order to improve performance, both CNNs are trained on related images using the Adam optimizer. The proposed method has been evaluated on the Herlev benchmark database and has provided 99.23 percent accuracy for the two-classes case and 97.65 percent accuracy for the 7-classes case. The results have shown that the presented method has provided higher accuracy than the baseline networks and many existing methods.

The rapid development of Internet information technology has made e-commerce enterprises face complex and changing financial problems. Combining artificial intelligence algorithms and dynamic monitoring of financial risks has been a current research hotspot. Based on this, this paper conducts an empirical study with a sample of listed Chinese e-commerce enterprises from 2012 to 2022. Firstly, using factor analysis (FA) to obtain the common factors between the original financial and non-financial indicators has the effect of reducing the overfitting risk of the model. Secondly, the mean square error (MSE) of the output and predicted values of the Long Short-Term Memory neural network (LSTM) is used as the fitness function of the intelligent swarm optimization algorithm, and then the Particle Swarm Optimization (PSO) algorithm is used to optimize the learning rate (LR) and the number of hidden layer neurons in the Long Short-Term Memory (LSTM) neural network. Finally, a financial risk prediction model based on FA-PSO-LSTM deep learning is constructed, and multiple benchmark models are introduced for comparative analysis on each evaluation index. The study shows that for nonlinear multivariate data with complex structure, the fused deep learning model proposed in this paper achieves the lowest values in mean square error (MSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). This indicates that the model has the best prediction effect, which is helpful to help managers make relevant decisions efficiently and scientifically and make the enterprise sustainable.

An optimized neural network modeling approach with genetic algorithm for short-term load forecasting based on only multiple delayed historical power load data is proposed. Genetic algorithm is used to globally optimize the number of delayed input neurons and the number of neurons in the hidden layer of the neural network architecture. Modification of Levenberg-Marquardt algorithm with Bayesian regularization is used to improve the generalization ability of the neural network. The performance of our proposed approach has been compared using actual power load data sets. Numerical results show that our proposed power load forecasting approach is comparable to the existing approaches that use multiple input variables such as power load data, day type load patterns and weather conditions

This paper investigates the error of determining the coordinates of an acoustic signal source depending on various parameters of the developed software-mathematical models for training a neural network. The authors determined the optimal values of these parameters according to the criterion of minimum error. The study found that the optimal amount of data for training a neural network is 25,000 training pairs, and the number of hidden layers should not exceed 5. Among several forms of sensor arrangement, the best is the square, as it provides a smaller error and uniform filling of the microphone’s perimeter. Their number should not exceed 9, since the best algorithm for training the neural network trainlm requires significant computing resources of the system on which training is performed. An increase in the number of sensors leads to an increase in the training time of the neural network but does not lead to an improvement in accuracy. When studying the effect of the distance to the acoustic signal source and the algorithm for selecting the base sensor for calculating the time difference, it was found that changing these parameters does not affect the accuracy of the neural network. The best structure of the number of hidden neurons in each hidden layer is an expanding structure, in which the number of neurons in each layer increases from the input layer to the middle-hidden layer, then decreases to the output layer. Using the optimized parameters, the neural network was trained and its accuracy was evaluated on two data sets: those used for training and those unknown to the neural network. In both cases, the neural network demonstrated good results. According to the results of the study, it was found that the error of the optimised system is 6.7681 × 10-5 m in the Y coordinate and 6.6120 × 10-5 m in the X coordinate.

In view of the inconvenience of parameter setting in the existing fault location method based on deep neural network, this paper combines the global random search ability of Genetic Algorithm, L2 regularization to prevent model overfitting and deep neural network to learn complex nonlinear ability, a fault location algorithm based on G-DNN is proposed. The optimal number of hidden layer neurons, learning rate, training period and L2 regularization coefficient of deep neural network are obtained by Genetic Algorithm; Input the statement coverage information and state values into the deep neural network to calculate the suspiciousness value of each executable statement; Locate defects from high to low according to the suspiciousness values. The Siemens Suite is used as the experimental sample, and the G-DNN is compared with the five defect location algorithms. The results show that the defect of the G-DNN can locate defects more accurately, and the positioning efficiency is improved.

Based on the characteristics of remote sensing images of mine vegetation, this research studied the application of deep belief network model in mine vegetation identification. Through vegetation identification and classification, the ecological environment index of mining area was determined according to the analysis of vegetation and coverage. Deep learning algorithm is adopted to improve the depth study, the vegetation coverage in the analysis was studied. Parameters and parameter values were selected for identification by establishing the optimal experimental design. The experimental results were compared with remote sensing images to determine the accuracy of deep learning identification and the effectiveness of the algorithm. When the sample size is 2,000,000 pixels, through repeated tests and classification effect comparison, the optimal parameter setting suitable for mine vegetation identification is obtained. Parameter setting: the number of network layers is 3 layers; the number of hidden layer neurons is 60. The learning rate is 0.01 and the number of iterations is 2. The average recognition rate of vegetation coverage was 95.95%, outperforming some other models, and the accuracy rate of kappa coefficient was 0.95, which can accurately reflect the vegetation coverage. The clearer the satellite image is, the more accurate the recognition result is, and the accuracy is closer to 100%. The identification of vegetation coverage has important guiding significance for determining the area and area of ecological restoration.

Tire defects are crucial for safe driving. Specialized experts or expensive tools such as stereo depth cameras and depth gages are usually used to investigate these defects. In image processing, feature extraction, reduction, and classification are presented as three challenging and symmetric ways to affect the performance of machine learning models. This paper proposes a hybrid system for cracked tire detection based on the adaptive selection of correlation features and deep belief neural networks. The proposed system has three steps: feature extraction, selection, and classification. First, the oriented gradient histogram extracts features from the tire images. Second, the proposed adaptive correlation feature selection selects important features with a threshold value adapted to the nature of the images. The last step of the system is to predict the image category based on the deep belief neural networks technique. The proposed model is tested and evaluated using real images of cracked and normal tires. The experimental results show that the proposed solution performs better than the current studies in effectively classifying tire defect images. The proposed hybrid cracked tire detection system based on adaptive correlation feature selection and Deep Belief Neural Networks’ performance provided better classification accuracy (88.90%) than that of Belief Neural Networks (81.6%) and Convolution Neural Networks (85.59%).

Artificial neural networks (ANNs) especially Backpropagation Neural Network (BPNN) was used largely in image processing. The backpropagation neural network algorithm (BP) was used for training the BPNN for image compression/decompression. The BP requires long time to train the BPNN with small error. Therefore, in this research, a three layered BPNN was designed for building image compression system. The Fast backpropagation neural network algorithm (FBP) was used for training the designed BPNN to reduce the training time (convergence time) of BPNN as possible as. Many techniques were used to improve the use of FBP for BPNN training. This is done by using different architecture of BPNN by changing the number of input layer neurons and number of hidden layer neurons. Also we trained the BPNN with different FBP parameters. Finally, FBP results such as compression ratio (CR) and peak signal to noise ratio (PSNR) are computed and compared with BP results. From the results, we noticed that the use of FBP improve the BPNN training by reducing the convergence time of image compression learning process.

1 - Congruence of deep learning in biomedical engineering: future prospects and challenges

A large database needs a heavy computation when the analysis is needed. The heavy computation leads to decrease the autonomous system performance. In our previous work, a complete vision based dirvable road detection method was proposed using Deep Belief Neural Network(DBNN). However, the previous method is unable to perform in real time for a large scale database. Due to solve this problem, in this paper, two fast drivable road detection approaches have been proposed using Kernel Principal Component Analysis-Deep Belief Neural Network (KPCA-DBNN) and Dimensionality Reduction Deep Belief Neural Network (DRDBNN) to reduce heavy computation for a large database. In the KPCA-DBNN, KPCA is used for dimensionality reduction and DBNN is used for classification. In the DRDBNN, two DBNNs are used. One DBNN is used for dimensionality reduction, and other DBNN is used for classification. The performance of the two approaches is demonstrated by the experimental results. From the experimental results, we see that the KPCA-DBNN and DRDBNN approaches reduce the processing time as compared to the conventional DBNN method. In addition, the results indicate that DRDBNN performed better than KPCA-DBNN in terms of detection accuracy on a large road database.