And Neural Networks Research Articles

A Machine Learning Approach to Predict Channel Stability Using Channel Quality Index

Channel quality feedback is crucial for the operation of 4G and 5G radio networks, as it allows to control User Equipment (UE) connectivity, transmission scheduling, and the modulation and rate of the data transmitted over the wireless link. However, when such feedback is frequent and the number of UEs in a cell is large, the channel may be overloaded by signalling messages, resulting in lower throughput and data loss. Optimizing this signaling process thus represents a key challenge. In this paper, a focus on Channel Quality Indicator (CQI) reports that are periodically sent from a UE to the base station, and propose mechanisms to optimize the reporting process with the aim of reducing signaling overhead and avoiding the associated channel overloads, particularly when channel conditions are stable. To this end, we apply machine learning mechanisms to predict channel stability, which can be used to decide if the CQI of a UE is necessary to be reported, and in turn to control the reporting frequency. For this purpose,four machine learning models,namely Support Vector Machines (SVM),K-Nearest Neighbour (KNN),Random Forest and Neural Networks (ANN). Simulation results shows that both provide a high prediction accuracy, with ANN consistently outperforming, especially as CQI reporting frequency reduces. Keywords: ANN,KNN,SVM,Random Forest,CQI,5G,UE

Predicting Hydraulic Jump Length on Rough Beds Using Data-Driven Models

The hydraulic jump can be used for some purpose such as dissipating the flow energy in order to prevent bed erosion; aerating water and facilitating the mixing procedure of chemical that used for water purification. In this paper, various artificial intelligence (AI) models including gene expression programming (GEP), adaptive-neuro-fuzzy inference system with grid partition (ANFIS-GP), and neural networks (ANNs) were used to estimate developed and non-developed hydraulic jump length. Four various GEP, ANFIS-GP and ANN models including different combinations of Froude number, bed roughness height, upstream and downstream flow depth based on measured experimental data-set were developed to estimate hydraulic jump length variations. The root mean squared error (RMSE) and determination coefficient (R2) indices were applied for testing models’ accuracy. Regarding the comparison results, it was seen that the ANFIS-GP, ANN, and GEP models could be employed successfully in estimating hydraulic jump length. The comparison between three AI approaches emphasized the superiority of ANNs and ANFIS-GP over the other intelligent models for modeling developed and non-developed hydraulic jump length, respectively. For non-developed hydraulic jump, the R2 and RMSE values obtained as 0.87 and 2.84 for ANFIS-GP model.

Screening of Mangifera indica L. functional content using PCA and neural networks (ANN)

A method using reversed-phase high-performance liquid chromatography with diode array detection (HPLC-DAD) was applied after extraction with acidified methanol, to determine 12 bioactive phenolic compounds in the peel and pulp of the mango fruit (Mangifera indica L.) cultivated in the Bahia, state of Brazil. The proposed methodology was previously fully validated and proven successful in the analysis of methanolic extracts of lyophilized samples. The limits of quantification ranged between 0.78 and 3.14 mg L−1 and the individual recovery values obtained for the spiked samples ranged from 80% to 120%. The results were evaluated using PCA and ANN. The results indicate that the fruits are rich in polyphenols, mainly: ellagic acid, gallic acid, rutin and catechin, which contribute to their greater use as functional foods, natural antioxidants and in the pharmaceutical industry, as well as other applications.

Improving Land Use/Cover Classification with a Multiple Classifier System Using AdaBoost Integration Technique

Guangzhou has experienced a rapid urbanization since 1978 when China initiated the economic reform, resulting in significant land use/cover changes (LUC). To produce a time series of accurate LUC dataset that can be used to study urbanization and its impacts, Landsat imagery was used to map LUC changes in Guangzhou from 1987 to 2015 at a three-year interval using a multiple classifier system (MCS). The system was based on a weighted vector to combine base classifiers of different classification algorithms, and was improved using the AdaBoost technique. The new classification method used support vector machines (SVM), C4.5 decision tree, and neural networks (ANN) as the training algorithms of the base classifiers, and produced higher overall classification accuracy (88.12%) and Kappa coefficient (0.87) than each base classifier did. The results of the experiment showed that, based on the accuracy improvement of each class, the overall accuracy was improved effectively, which combined advantages from each base classifier. The new method is of high robustness and low risk of overfitting, and is reliable and accurate, and could be used for analyzing urbanization processes and its impacts.

Exergo-ecological evaluation of heat exchanger

Thermodynamic optimization of thermal devices requires information about the influence of operational and structural parameters on its behaviour. The interconnections among parameters can be estimated by tools such as CFD, experimental statistic of the deviceetc. Despite precise and comprehensive results obtained by CFD, the time of computations is relatively long. This disadvantage often cannot be accepted in case of optimization as well as online control of thermal devices. As opposed to CFD the neural network or regression is characterized by short computational time, but does not take into account any physical phenomena occurring in the considered process. The CFD model of heat exchanger was built using commercial package Fluent/Ansys. The empirical model of heat exchanger has been assessed by regression and neural networks based on the set of pseudo-measurements generated by the exact CFD model. In the paper, the usage of the developed empirical model of heat exchanger for the minimisation of TEC is presented. The optimisationconcerns operational parameters of heat exchanger. The TEC expresses the cumulative exergy consumption of non-renewable resources. The minimization of the TEC is based on the objective function formulated by Szargut. However, the authors extended the classical TEC by the introduction of the exergy bonus theory proposed by Valero. The TEC objective function fulfils the rules of life cycle analysis because it contains the investment expenditures (measured by the cumulative exergy consumption of non-renewable natural resources), the operation of devices and the final effects of decommissioning the installation.

Performance of models based on a linear regression and neural networks

In this paper the comparison of models based on a linear regression and neural networks is presented. The analyzed models are the generalized profile function models, GPFM. The GPFM provides approximations of the individual models (individual stem profile models) of the objects using only two basic measurements. The performances of the obtained GPFM, by using the linear regression relations and neural networks are compared by a test platform in MATLwith a simple graphic user interface. It is shown that application of both linear regression and neural networks provides the efficient and robust generalized model with very good performances.

Structural Damage Identification Using Wavelet Packet Analysis and Neural Network

This paper describes a study on damage identification using wavelet packet analysis and neural networks. The identification procedure could be divided into three steps. First, structure responses are decomposed into wavelet packet components. Then, the component energies are used to define damage feature and to train neural network models. Finally, in combination with the feature of the damaged structure response, the trained models are employed to determine the occurrence, the location and the qualification of the damage. The emphasis of this study is put on multi-damage case. Relevant issues are studied in detail especially the selection of training samples for multi-damage identification oriented neural network training. A frame model is utilized in the simulation cases to study the sampling techniques and the multi-damage identification. Uniform design is determined to be the most suitable sampling technique through simulation results. Identifications of multi-damage cases of the frame including different levels of damage at various locations are investigated. The results show that damages are successfully identified in all cases.