查看更多>>摘要：This article proposes a VGG network with histogram of oriented gradient(HOG)feature fusion(HOG-VGG)for polarization synthetic aperture radar(PolSAR)image terrain classification.VGG-Net has a strong ability of deep feature extraction,which can fully extract the global deep features of different terrains in PolSAR images,so it is widely used in PolSAR terrain classification.However,VGG-Net ignores the local edge & shape features,resulting in incomplete feature representation of the PolSAR terrains,as a consequence,the terrain classification accuracy is not promising.In fact,edge and shape features play an important role in PolSAR terrain classification.To solve this problem,a new VGG network with HOG feature fusion was specifically proposed for high-precision PolSAR terrain classification.HOG-VGG extracts both the global deep semantic features and the local edge & shape features of the PolSAR terrains,so the terrain feature representation completeness is greatly elevated.Moreover,HOG-VGG optimally fuses the global deep features and the local edge & shape features to achieve the best classification results.The superiority of HOG-VGG is verified on the Flevoland,San Francisco and Oberpfaffenhofen datasets.Experiments show that the proposed HOG-VGG achieves much better PolSAR terrain classification performance,with overall accuracies of 97.54％,94.63％,and 96.07％,respectively.
查看更多>>摘要：To gain a more comprehensive understanding and evaluate foam aluminum's performance,researchers have introduced various characterization indicators.However,the current understanding of the significance of these indicators in analyzing foam aluminum's performance is limited.This study employs the Generalized Regression Neural Network(GRNN)method to establish a model that links foam aluminum's microstructure characterization data with its mechanical properties.Through the GRNN model,researchers extracted four of the most crucial features and their corresponding weight values from the 13 pore characteristics of foam aluminum.Subsequently,a new characterization formula,called"Wang equivalent porosity"(WEP),was developed by using residual weights assigned to the feature weights,and four parameter coefficients were obtained.This formula aims to represent the relationship between foam aluminum's microstructural features and its mechanical performance.Furthermore,the researchers conducted model verification using compression data from 11 sets of foam aluminum.The validation results showed that among these 11 foam aluminum datasets,the Gibson-Ashby formula yielded anomalous results in two cases,whereas WEP exhibited exceptional stability without any anomalies.In comparison to the Gibson-Ashby formula,WEP demonstrated an 18.18％improvement in evaluation accuracy.
查看更多>>摘要：This study considers a superconducting electrodynamic maglev train of MLX01 type.The characteristics of the electromagnetic spring coefficient of a single bogie under different magnetomotive force(MF)of the superconducting coil and standard air gap(Sag)were explored.In view of the small electromagnetic damping,a passive damping control strategy and an active damping control strategy were designed to increase the electromagnetic damping force between the superconducting coil and ground coil.Combined with the coupling numerical model of a single bogie,the vibration characteristics of the bogie in different directions with different damping control strategies were studied when the Sag and MF were fixed.The results can provide important theoretical support for stable operation control of maglev trains.
查看更多>>摘要：Axiomatization of Shannon entropy is a subject that has received lots of attention in the information theory literature.While Shannon entropy is defined on probability distribution,we define a new type of entropy on the set of partitions of finite subsets of metric spaces,which has a rich algebraic structure as a partially ordered set.We propose an axiomatization of an entropy-like measure of partitions of sets of objects located in metric spaces,and we derive an analytic expression of this new type of entropy referred to as inertial entropy.This approach starts with the notion of inertia of a partition and includes a study of the behavior of the sum of square errors of a partition.In this context,we characterize the chain of partitions produced by the Ward hierarchical clustering method.Starting from inertial entropies of partitions,we introduce conditional entropies which,in turn,generate metrics on partitions of finite sets.These metrics are used as external validation tools for clusterings of labeled data sets.The metric generated by inertial entropy can be used to validate data clustering for labeled data sets.This type of validation aims to determine to what extend labeling of the data coincides with the clustering obtained algorithmically,and we obtain a high degree of consistency of the data labeling with the results of several hierarchical clusterings.
查看更多>>摘要：Network attack detection and mitigation require packet collection,pre-processing,feature analysis,classification,and post-processing.Models for these tasks sometimes become complex or inefficient when applied to real-time data samples.To mitigate hybrid assaults,this study designs an efficient forensic layer employing deep learning pattern analysis and multidomain feature extraction.In this paper,we provide a novel multidomain feature extraction method using Fourier,Z,Laplace,Discrete Cosine Transform(DCT),1D Haar Wavelet,Gabor,and Convolutional Operations.Evolutionary method dragon fly optimisation reduces feature dimensionality and improves feature selection accuracy.The selected features are fed into VGGNet and GoogLeNet models using binary cascaded neural networks to analyse network traffic patterns,detect anomalies,and warn network administrators.The suggested model tackles the inadequacies of existing approaches to hybrid threats,which are growing more common and challenge conventional security measures.Our model integrates multidomain feature extraction,deep learning pattern analysis,and the forensic layer to improve intrusion detection and prevention systems.In diverse attack scenarios,our technique has 3.5％higher accuracy,4.3％higher precision,8.5％higher recall,and 2.9％lower delay than previous models.
查看更多>>摘要：Location prediction in social media,a growing research field,employs machine learning to identify users'locations from their online activities.This technology,useful in targeted advertising and urban planning,relies on natural language processing to analyze social media content and understand the temporal dynamics and structures of social networks.A key application is predicting a Twitter user's location from their tweets,which can be challenging due to the short and unstructured nature of tweet text.To address this challenge,the research introduces a novel machine learning model called the location-aware attention LSTM(LAA-LSTM).This hybrid model combines a Long Short-Term Memory(LSTM)network with an attention mechanism.The LSTM is trained on a dataset of tweets,and the attention network focuses on extracting features related to latitude and longitude,which are crucial for pinpointing the location of a user's tweet.The result analysis shows approx.10％improvement in accuracy over other existing machine learning approaches.
Salman Khayoon AldriasawiNihayat Hussein AmeenKareem Idan FadheeAshham Muhammed Anead...
78-92页
查看更多>>摘要：The present study establishes a new estimation model using an artificial neural network(ANN)to predict the mechanical properties of the AISI 1035 alloy.The experiments were designed based on the L16 orthogonal array of the Taguchi method.A proposed numerical model for predicting the correlation of mechanical properties was supplemented with experimental data.The quenching process was conducted using a cooling medium called"nanofluids".Nanoparticles were dissolved in a liquid phase at various concentrations(0.5％,1％,2.5％,and 5％vf)to prepare the nanofluids.Experimental investigations were done to assess the impact of temperature,base fluid,volume fraction,and soaking time on the mechanical properties.The outcomes showed that all conditions led to a noticeable improvement in the alloy's hardness which reached 100％,the grain size was refined about 80％,and unwanted residual stresses were removed from 50 to 150 MPa.Adding 5％of CuO nanoparticles to oil led to the best grain size refinement,while adding 2.5％of Al2O3 nanoparticles to engine oil resulted in the greatest compressive residual stress.The experimental variables were used as the input data for the established numerical ANN model,and the mechanical properties were the output.Upwards of 99％of the training network's correlations seemed to be positive.The estimated result,nevertheless,matched the experimental dataset exactly.Thus,the ANN model is an effective tool for reflecting the effects of quenching conditions on the mechanical properties of AISI 1035.
万方数据
维普