Feed Forward Back Propagation Algorithm Research Articles

Artificial neural network modelling of faecal coliform removal in an intermittent cycle extended aeration system-sequential batch reactor based wastewater treatment plant

Artificial neural network (ANN) models have been designed to predict faecal coliform and total coliform removal for an intermittent cycle extended aeration-sequential batch reactor (ICEAS-SBR). Wastewater influent pH, biochemical oxygen demand (BOD), chemical oxygen demand (COD), total suspended solids (TSS), oil & grease (O&G), total kjeldahl nitrogen (TKN), ammonical nitrogen (AN), total phosphorus (TP), faecal coliform, and total coliform were used to develop the network. The data used to train and test the network were obtained from the Jamnagar municipal corporation-waste water treatment plant (JMC-WWTP). Feedforward backpropagation algorithm with learngdm learning function was used to develop ANN models. The number of neurons in hidden layer were varied between 2–10 to find the most reliable network. The optimum ANN models were selected for faecal coliform and total coliform on trial-and-error method. The performance of designed models was tested by computing correlation coefficient, mean absolute deviation (MAD), root mean square error (RMSE), and mean absolute percentage error (MAPE). The network with 6 hidden neurons was optimum for faecal coliform, and 8 hidden neurons was optimum for total coliform. The produced simulation results were within 5 % of MAPE for both faecal and total coliform. The ANN models allow faecal coliform and total coliform levels in the treated wastewater effluent to be regulated that reduces the public and in particular the oyster consumer’s health risks.

Retraction Note to: Prediction compressive strength of Portland cement-based geopolymers by artificial neural networks

In the present study, compressive strength results of geopolymers produced by ordinary Portland cement (OPC) as aluminosilicate source have been modeled by artificial neural networks. Six main factors including NaOH concentration, water glass to NaOH weight ratio, alkali activator to cement weight ratio, oven curing temperature, oven curing time and water curing regime each at 4 levels were considered for designing. A total of 32 experiments were conducted according to the L32 array proposed by the method. The neural network models were constructed by 10 input parameters including NaOH concentration, water glass to NaOH weight ratio, alkali activator to cement weight ratio, oven curing temperature, oven curing time, water curing regime, water glass content, NaOH content, Portland cement content and test trial number. The value for the output layer was the compressive strength. According to the input parameters in feed-forward back-propagation algorithm, the constructed networks were trained, validated and tested. The results indicate that artificial neural networks model is a powerful tool for predicting the compressive strength of the geopolymers in the considered range.

RETRACTED ARTICLE: Preserving mobile commerce IoT data using light weight SIMON block cipher cryptographic paradigm

Internet of Things (IoT) data security is one of the critical ideas in data security as the data to be transferred should be made secure. In the field of mobile commerce is the high effect that it will have on a few parts of the regular life and behavior of potential clients. In this present examination, we focused to build up the IoT security in the field of mobile commerce utilizing new cryptographic strategies. Fundamentally the sensitive data are classified from the entire dataset to enhance the accuracy utilizing Feed Forward Back Propagation Algorithm (FFBN). At that point the security of sensitive data is upgraded by utilizing Light Weight Cryptography (LWC) which encodes the input sensitive data called encryption. To enhance the data privacy and secrecy, an imaginative security model is proposed for example Lightweight SIMON block cipher. This will encrypt the data alongside optimal key selection; this LWC upgrades the m commerce data security level in cloud. For key optimization, a meta-heuristic algorithm called Crow Search Algorithm (CSA) is introduced. The proposed SIMON-CSA accomplishes least time to generate key value to decrypt the data.

Prediction of maximum pressure of journal bearing using ANN with multiple input parameters

ABSTRACT The purpose of this paper is to predict the maximum pressure of journal bearing using ANN. The maximum pressure values are analysed using FEM and predicted using ANN. The FEM analysis is performed for micropolar lubricated hybrid journal bearing, and the results are used for training and testing the ANN model. Feedforward backpropagation algorithm is used for ANN model development. The externally applied load and rotational speed of the journal are considered as input parameters for performance predictions. ANN predictions are made within and outside the prescribed range of input parameters. This approach shows better predictions for journal bearing performance. Results obtained from FEM and ANN are in close agreement with each other. The percentage error less than 0.5 is observed for training and testing of the ANN model. The prediction error is in the range of −0.6% to 0.48%. In this paper, a mathematical model is established using FEM. The performance predictions obtained using ANN are very useful because much time can be saved which would be otherwise consumed in the experimental or theoretical bearing analysis.

Remote Sensing of Time-Varying Tidal Flat Topography, Jiangsu Coast, China, Based on the Waterline Method and an Artificial Neural Network Model

Measurement of beach heights in the intertidal zone has great importance for dynamic geomorphology research, coastal zone management, and the protection of ecological resources. Based on satellite images, the waterline method based on satellite images is one of the most effective methods for constructing digital elevation models (DEMs) for large-scale tidal flats. However, for fast-changing areas, such as Tiaozini in the Jiangsu coast, timely and detailed topographical data are difficult to obtain due to the insufficient images over a short period of time. In this study, as a supplement to the waterline method, an artificial neural network (ANN) model with the multi-layer feed-forward back propagation algorithm was developed to simulate the topography of variable Tiaozini tidal flats. The “7-15-15-1” double hidden layers with optimized training structures were confirmed via continuous training and comparisons. The input parameters included spectral bands (HJ-1 images B1~B4), geographical coordinates (X, Y), and the distance (D) to waterlines, and the output parameter was the elevation. The model training data were the HJ-1 image for 21 March 2014, and the corresponding topographic data obtained from the waterline method. Then, this ANN model was used to simulate synchronous DEMs corresponding to remote sensing images on 11 February 2012, and 11 July 2013, under low tide conditions. The height accuracy (root mean square error) of the two DEMs was about 0.3–0.4 m based on three transects of the in-situ measured data, and the horizontal accuracy was 30 m—the same as the spatial resolution of the HJ-1 image. Although its vertical accuracy is not very high, this ANN model can quickly provide the basic geomorphological framework for tidal flats based on only one image. This model, therefore, provides an effective way to monitor rapidly changing tidal flats.

Fractal and multifractional-based predictive optimization model for stroke subtypes’ classification

Numerous natural phenomena display repeating self-similar patterns. Fractal is used when a pattern seems to repeat itself. Fractal and multifractal methods have extensive applications in neurosciences in which the prevalence of fractal properties like self-similarity in the brain, equipped with a complex structure, in medical data analysis at various levels of observation is admitted. The methods come to the fore since subtle details are not always detected by physicians, but these are critical particularly in neurological diseases like stroke which may be life-threatening. The aim of this paper is to identify the self-similar, significant and efficient attributes to achieve high classification accuracy rates for stroke subtypes. Accordingly, two approaches were implemented. The first approach is concerned with application of the fractal and multifractal methods on the stroke dataset in order to identify the regular, self-similar, efficient and significant attributes from the dataset, with these steps: a) application of Box-counting dimension generated BC_stroke dataset b) application of Wavelet transform modulus maxima generated WTMM_stroke dataset. The second approach involves the application of Feed Forward Back Propagation (FFBP) for stroke subtype classification with these steps: (i) FFBP algorithm was applied on the stroke dataset, BC_stroke dataset and WTMM_stroke dataset. (ii) Comparative analyses were performed based on accuracy, sensitivity and specificity for the three datasets. The main contribution is that the study has obtained the identification of self-similar, regular and significant attributes from the stroke subtypes datasets by following multifarious and integrated methodology. The study methodology is based on the singularity spectrum which provides a value concerning how fractal a set of points are in the datasets (BC_stroke dataset and WTMM_stroke dataset). The experimental results reveal the applicability, reliability and accuracy of our proposed integrated method. No earlier work exists in the literature with the relevant stroke datasets and the methods employed. Therefore, the study aims at pointing a new direction in the relevant fields concerning the complex dynamic systems and structures which display multifractional nature.

Screening Reservoir Candidates for Enhanced Oil Recovery (EOR) in Angolan Offshore Projects

The neuro-fuzzy (NF) approach presented in this work is based on five (5) layered feedforward backpropagation algorithm applied for technical screening of enhanced oil recovery (EOR) methods. Associated reservoir rock-fluid oilfield data from successful EOR projects were used as input and predicted output in the training and validation processes, respectively. The developed model was then tested by using data set from Block B of an Angolan oilfield. The results of the sensitivity analysis between the Mamdani and the Takagi-Sugeno-Kang (TSK) approach incorporated in the algorithm has shown the robustness of the TSK ANFIS (Adaptive Neuro-Fuzzy Inference System) approach in comparison to the other approach for the prediction of a suitable EOR technique. The simulation test results showed that the model presented in this study can be used for technical selection of suitable EOR techniques. Within the area investigated (Block B, Angola) polymer, hydrocarbon gas, and combustion were identified as the suitable techniques for EOR.

Design and Implementation of Intelligent Hybrid Neuro Controller for Automatic Voltage Regulator

In this paper, we presented a hybrid controller by combining the advantages of both PID and Neuro controllers for automatic voltage regulator (AVR) system. A Neuro controller is designed using multilayer feedforward neural network and Levenberg-Marquardt backpropagation algorithm is used for training the network. Also, hybrid controller is achieved by blending the characteristics of classical PID and proposed Neuro controller using a switching mechanism based on the error. The proposed PID, Neuro and Hybrid controllers are simulated in MATLAB environment and their transient response parameters are compared. The simulation results clearly indicated the improvement in the transient output of the automatic voltage regulator system with proposed hybrid controller even in the presence of uncertainties in the system.

Classification of Urine Odour Using Artificial Neural Networks

Introduction An odour create a distinctive smell that is caused by specific volatile molecules; it can be either a single chemical compound or a mixture of them. Volatile organic compounds are currently of great interest for non-invasive disease diagnoses such as diabetes [2] and prostate cancers [3]. In this work, the detection of human urine is studied in order to aid incontinent individuals and prevent social embarrassment. Urine odour is known to comprise of more than 200 different types of chemical compounds, including ketones, aldehydes, and esters [1]. Gas Sensing Device The proposed sensing device is compact with three custom-made CMOS non-specific metal oxide semiconductors coated with n-type tungsten trioxide, Pd/Pt doped tin oxide, and p-type copper oxide, respectively. The sensors are interfaced to an integrated Teensy 3.6 microcontroller; and the response data are collected at a 10 Hz sampling rate. This device has been tested via a custom gas testing station with synthetic urine as well as its three individual interference compounds, ammonia, ethyl acetate and acetone. The maximum concentration levels tested in air were 25 ppm, 20 ppm and 200 ppm, respectively, and 5% for synthetic urine. Five repetitions were performed with five concentration steps and three different humidity levels for each gas of interest. The three sensors responded to the target compounds differently with a specific pattern that can be classified through pre-processing and supervised artificial neural networks. Artificial Neural Networks The collected time-series data were first pre-processed to extract the change of sensor response, and auto-ranged to spread the sample variance evenly across all sensors. This step allows the input sensor data to contribute equally to the neural network models and increase the learning speed and prediction accuracy. Two methods were trialled, a shallow neural network with multi-layer perceptrons, and a deep convolutional neural network (CNN). The shallow neural network topology consists of 4 input neurons (no concentration information), 2 hidden layers of 8 neurons each and 4 output neurons for each odour class. This supervised feed-forward backpropagation algorithm is based on the Levernberg-Marquardt approach with sigmoid function for hidden layers. The convolutional neural network has two convolution layers and two pooling layers before the fully connected output layer, which works in the same manner as a shallow neural network. Unlike a typical 2D CNN used in image processing, this 1D network has the kernel size of 1 × n and a single row filtering window tiling down the dataset. In both models, four experiment repeats were used for training and one for testing, until all had been independently tested. Results and Conclusions Confusion matrices are used to demonstrate the odour prediction result of both models, as presented in Table 1 for the shallow neural network and Table 2 for the convolutional neural network. These tables are the two model results for the same dataset. Two performance evaluation methods are studied here, the cross entropy and the precision. The cross entropy loss is used to measure the model performance by comparing the predicted outputs with the target outputs. It is calculated through a log equation. For a classification model, cross entropy of 0 indicates perfect results. In the shallow neural network model, the cross entropy is 0.0269 for Table 1, and 0.0183 for Table 2 with the CNN model. Precision is defined as the accuracy of positive predictions and calculated on a per-class basis. The precision of urine odour prediction is 0.93 for the shallow neural network and 1.0 for the CNN, which is ideal. Therefore, while both have high overall accuracies, 92.56% for the shallow neural network and 97.22% for the CNN, the CNN provides better results not just in the urine odour detection, but also the predictions of other similar compounds. The shallow neural network has its merits in terms of easy implementation of a real-time classification device for incontinence individuals, and low memory requirement with relatively high accuracy. Overall, this work proves the feasibility of using these low-cost metal oxide sensors for urine prediction in an open environment that could contain interfering VOCs such as solvents.

Deep neural network based prediction of burst parameters for Zircaloy-4 fuel cladding during loss-of-coolant accident

BackgroundUnderstanding the behaviour of nuclear fuel claddings by conducting burst test on single cladding tube under simulated loss-of-coolant accident conditions and developing theoretical cum empirical predictive computer codes have been the focus of several investigations. The developed burst criterion (a) assumes symmetrical deformation of cladding tube in contrast to experimental observation (b) interpolates the properties of Zircaloy-4 cladding in mixed α+ β phase (c) does not account for azimuthal temperature variations. In order to overcome all these drawbacks of burst criterion, it is reasoned that artificial intelligence technique may be a better option to predict the burst parameters. MethodsArtificial neural network models based on feedforward backpropagation algorithm with logsig transfer function are developed. ResultsNeural network architecture of 2-4-4-3, that is model with two hidden layers having four nodes in each layer is found to be the most suitable. The mean, maximum, and minimum prediction errors for this optimised model are 0.82%, 19.62%, and 0.004%, respectively. ConclusionThe burst stress, burst temperature, and burst strain obtained from burst criterion have average deviation of 19%, 12%, and 53% respectively whereas the developed neural network model predicted these parameters with average deviation of 6%, 2%, and 8%, respectively.

Feed-forward back-propagation (FFBP) algorithm for property prediction in friction stir spot welding of aluminium alloy

Facing the issue of cost and efficiency in experiments and tests in determining the properties of the welded structure is a challenge in friction stir spot welding (FSSW) optimization. Employing the machine learning technique of artificial neural network (ANNs) to develop a prediction model with fewer experiments and tests is a gentle solution to forecast the properties of the spot weld structures. In this study, the extended full factorial design with respect to thetool speed, plunge depth, anddwell time are applied to the FSSW specimens of aluminium A5052-H122 of 2mm thick through 27 experiments and evaluated via tensile shearing load testing. The multilayer neural network of feed-forward and back-propagation (FFBP) algorithm was engaged to learn and train the neural network iteratively with a set of weights and bias of 27 variations of inputs to fit the predicted tensile shear loads of the spot weld structures. Based on the resulted of regression plot, it is shown that the correlation coefficient (R) is perfect for training with the value of 0.999 and for testing the correlation coefficient (R) is reached to 0.958. However, the correlation coefficient is relatively good for validation with R equal to 0.921. For all data sets, the correlation coefficient is good with R of 0.833. It can be seen that the ANNs prediction model is relatively good since the correlation coefficient relatively close to 1.

Assessment of Suspended Sediment Load with Neural Networks in Arid Watershed

Here, the assessment of suspended sediment load is evaluated by four ANN algorithms: support vector machine (SVM), cascade-forward back-propagation (CFBP), feed-forward back-propagation (FFBP), and radial basis fewer neuron (RBFN) networks. Techniques are applied to a watershed of arid region, India. Sensitivity in terms of Nash–Sutcliffe coefficient (ENS), correlation coefficient (CC), and ratio between root mean square error and standard deviation (RSR) are computed. Results show that SVM shows preeminent value of RSR 0.0636, ENS 0.8869, and CC 0.9418, while Qt, Qt−1, Qt−2, Qt−3, Qt−4, Pt, Pt−1, Pt−2, Pt−3 architecture is applied. But for the same architecture, FFBP, RBFN and CFBP illustrate that the paramount value of CC is 0.9350, 0.9228, and 0.8985. As a whole, the performance of SVM shows superiority while considering various combinations of discharge and rainfall in contrast to FFBP, CFBP, and RBFN algorithm. Among all techniques, RBFN performs poor as compared to other algorithms. Interpretation of the results will help to compute sediment load in un-gauged catchments.

Modeling of glass fiber reinforced composites for optimal mechanical properties using teaching learning based optimization and artificial neural networks

The present work is aimed at determining mechanical properties of chopped strand glass fiber reinforced composite laminates manufactured based on the design of experiments by resin transfer molding at various injection pressures with 4, 5 and 6 layers. Response surface methodology was implemented to the experimental data for evaluating the effect of number of layers and resin injection pressure on mechanical properties and void content. Teaching learning based optimization (TLBO) has been proposed to predict optimal (maximum) mechanical properties of composite by optimizing the number of layers and injection pressure. Artificial neural network (ANN) with feed forward back propagation algorithm was also used to predict the responses and compare with experimental and TLBO results. It was found that the predicted values of responses from TLBO and ANN are good in agreement with experimental results.

Characterizing Complexity and Self-Similarity Based on Fractal and Entropy Analyses for Stock Market Forecast Modelling

Complex systems constitute components that interact with one another and involve phenomena which are not always easy to understand in terms of their components and interactions. Alternative mathematical models have been developed so that the users’ tasks can be facilitated and an actual assistance can be provided for decision-making processes in case of every encountered incident which requires critical decision-making. Within this framework, financial systems can be regarded as complex systems with their volatile and vulnerable nature along with various parameters and interactions involved. Forecasting shifts in stock indices is crucial to validate the potential strategies of monetary mechanisms. Therefore, forecasting is an essential step in financial decision-making to manage data selection and attain robust prediction. Our purpose is to optimize the stock indices’ forecasting model in the stock indices dataset, constructed from the daily values. The following steps were applied to demonstrate the critical significance of Hurst exponent (HE) computed by Rescaled Range (R/S) fractal analysis when used as indicator in conjunction with Shannon entropy (SE) and Renyi entropy (RE) for the future forecasting ability of the stock indices. With this aim, the following stages were performed with indicators obtained from the applications and added into the dataset in the respective order. The first stage consists of: i) HE indicator, ii) Entropy based SE and RE indicators, iii) HE, SE and RE indicators. As the second stage, stock indices day-to-day valuation was evaluated using Multi Layer Regression (MLR), Support Vector Regression (SVR) and Feed Forward Back Propagation (FFBP) algorithms, applied for each indicator for comparative analysis. When compared with earlier works, no relevant work exists in the literature in which the algorithms and above-mentioned indicators have been used in conjunction with one another. This paper, through the multistage methodology and proposed model, demonstrates that HE is obviously a significant and critical determining indicator compared to RE and SE indicators for forecasting purposes. Consequently, experimental results demonstrate the accuracy and applicability of the proposed method. Thus, this study attempts to illustrate a new frontier in domains concerning critical decision-making processes in non-linear, dynamic, and volatile environments.

Comparative study on the effects of meteorological and pollutant parameters on ANN modelling for prediction of SO2

Variations in meteorological parameters, different transportation mechanisms, complex reaction mechanisms and insufficient control measures have made control of ambient air pollution, a challenge. Prediction mechanisms for pollutant concentrations in advance have become necessary to regulate parameters within the acceptable limits. Artificial neural network (ANN) modelling can be used for predicting the concentrations of pollutants by establishing functional relationships between complex and nonlinear predictor variables and outputs. The application of time-series ANN models to predict air quality parameters was investigated by using six pollutant variables, five meteorological parameters and three time parameters to predict the concentration of SO2. An industrial belt in the southern part of India was selected as the study area. Two years of input parameters and feed-forward back propagation algorithm were used to construct the ANN model. Input parameters were optimized using forward selection and backward elimination techniques. Mean squared error and coefficient of determination were used to evaluate the models. The developed model exhibited very promising performance evaluation characteristics with the best model resulting in MSE of 0.0115 and R2 value of 0.8979. The hybrid ANN models employing input parameter optimization techniques resulted in better performance characteristic values than conventional models. Any model exhibited a minimum reduction in MSE by 9% and 3% improvement in correlation. Pollutant parameters were found to affect the ANN models when compared to the meteorological parameters. Predicted values compared to the National Ambient Air Quality Standards were as low as 55% of the maximum allowable concentration in ambient air.

Using atmospheric inputs for Artificial Neural Networks to improve wind turbine power prediction

A robust machine learning methodology is used to generate a site-specific power-curve of a full-scale isolated wind turbine operating in an atmospheric boundary layer to drastically improve the power predictions, and, thus, the forecasting of the monthly energy production estimates. The study has important implication in measuring the financial feasibility of wind farms by improving the accuracy of monthly energy estimates. The significance of the study is that atmospheric stability and air-density are accounted in the power predictions of the wind turbine. Artificial Neural Networks (ANN) machine learning approach is used to generate multi-parameter input models to estimate the power produced by the wind turbine. The ANN model in this study uses Feed Forward Back Propagation (FFBP) algorithm. The power- and wind-data is obtained from a 2.5 MW wind turbine that has a Meteorological tower located 900 m Southwest of the wind turbine in Kirkwood, Iowa, USA. The study investigates the role of atmospheric boundary-layer metrics – Wind Speed, Density (a measure of stratification), Richardson Number, turbulence intensity, and wind shear as input parameters into the ANN model. The study investigates the influence of FFBP ANN hyper-parameters on the power prediction accuracy. Comparison of the FFBP ANN model to other power curve correction techniques demonstrated an improvement in the Mean Absolute Error (MAE) of 40% when compared to the density correction (the next closest). The five-parameter 4-layer FFBP ANN has an average energy production error of 0.4% for the nine months while the IEC this error is −3.7% and for the air density correction the error is −1.9%, respectively. Finally, the study determines the performance of the FFBP ANN model for different atmospheric stability regimes (Unstable, Stable, Strongly Stable, Strongly Unstable and Neutral) classified using two criterions - Richardson number and Turbulence intensity. The largest MAE occurs during the strongly stable regime of the atmospheric boundary layer for both criteria.

Analysis of Satisfaction Level with Servqual method using Artificial Neural Networks

This research aims to analyze the level of student satisfaction at a college in pattern recognition using an artificial neural network with feedforward backpropagation algorithms. Input on the network consists of 5 variables, namely Tangibles (X1), Reliability (X2), Responsiveness (X3), Assurance (X4) and Empathy (X5) which are the dimensions of the Servqual method. Data processing in the Servqual method will become target data on the network. Data is obtained through multi-item questions given to students. After some training trials, the best configuration architecture consists of 5 neurons in the input layer, one hidden layer with seven neurons and one neuron in the output layer. The activation function used by a Sigmoid Binary that has a range of 0 to 1. This architecture can recognize the pattern well in the 14th iteration with MSE of 0.000192. Data testing results obtained an accuracy rate of 91.07% with an error rate of 8.93%. So that it can be concluded that the Artificial Neural Network with feedforward backpropagation algorithm in the case of analyzing the level of student satisfaction on college services can recognize the pattern well.

Measurement of performance and emission distinctiveness of Aegle marmelos seed cake pyrolysis oil/diesel/TBHQ opus powered in a DI diesel engine using ANN and RSM

The present investigation focuses on Artificial neural network (ANN) and Response surface methodology (RSM) modelling of a CI (Compression ignition) engine powered by Aegle marmelos (AM) pyrolysis oil/diesel/Tert-butyl hydroxyl quinone antioxidant (TBHQ) blend as a test fuel to predict and optimize the engine behaviour. Bio-oil is derived from AM de-oiled seed cake in a fixed bed pyrolysis reactor at 600 °C under the heating rate of 30 °C/min. To obtain data for testing and training the suggested RSM and ANN models, a direct injection, single cylinder CI engine was fuelled with proposed test fuel 80% diesel + 20% AM bio-oil + 1000 ppm TBHQ (A20D80T). The A20D80T has been assessed for the combined effects of varying compression ratio (CR = 16:1–17.5:1) and engine load (W = 25%–100%) in variable compression ratio (VCR) diesel engine through experimental investigation and ANN prediction and RSM optimization techniques. Using the experimental data for training, an ANN replica was developed according to feed forward back propagation algorithm (FFBP). Multi-layer perception (MLP) network was used for non-linear mapping between the experimental and predicted values. Engine process parameters were accurately predicted by trained ANN. The optimal values of engine performance (brake specific fuel consumption (BSFC) = 0.33 kg/kWh and brake thermal efficiency (BTE) = 22.01%) and emission behaviour (carbon monoxide (CO) = 0.67%, hydro carbon (HC) = 244 ppm, carbon dioxide (CO2) = 8.33% and oxides of nitrogen (NOx) = 351 ppm) were obtained by RSM optimization. The compression ratio of 17.5:1 at peak load condition was found to be superior engine characteristics through experimental assessment and ANN, RSM models. In the predicted ANN model the mean absolute average error (MAAE) was 0.552% and optimized RSM model MAAE was 1.231%. The ANN and RSM models gave the average correlation coefficient (R) of 0.998 and average coefficient of a determination (R2) of 0.991 respectively. The experimental, ANN and RSM analysis results depict that A20D80T blend delivered the enhanced performance and better emission behaviours compared with neat diesel fuel (D).

A Novel Approach of Mathematical Theory of Shape and Neuro-Fuzzy Based Diagnostic Analysis of Cervical Cancer.

This study aims to detect the abnormal growth of tissue in cervix region for diagnosis of cervical cancer using Pap test of patients. The proposed methodology classifies cervical cancer for pattern recognition either benign or malignant stages using shape and neuro-fuzzy based diagnostic model. In this experiment, firstly the authors segment Pap smear images of cervical cells using fuzzy c-means clustering algorithm and shape theory to classify them according to the presence of abnormality of the cells. Secondly the features extraction process is performed in the part of nucleus and cytoplasm on the squamous and glandular cells and the authors used input variables such as cytoplasm area (CA), cytoplasm circularity (CC), nucleus area (NA), nucleus circularity (NC), nucleus-cytoplasm ratio (NCR), and maximum nucleus brightness (MNB) in fuzzy tools and used fuzzy rules to evaluate the cervical cancer risk status as an output variable. The proposed neuro-fuzzy network system was developed for early detection of cervical cancer. A neural network was trained with 15-Pap image datasets where Levenberg-Marquardt(LM) a feed-forward back-propagation algorithm was used to get the status of the cervical cancer. Out of 15 samples database, 11 data set for training, 2 data set for validation and 2 data set for test were used in the ANN classification system. The presented fuzzy expert system(FES) successfully identified the presence of cervical cancer in the Pap smear images using the extracted features and the use of neuro-fuzzy system(NFS) for the identification of cervical cancer at the early stages and achieve a satisfactory performance with 100% accuracy.

Regarding Solid Oxide Fuel Cells Simulation through Artificial Intelligence: A Neural Networks Application

Because of their fuel flexibility, Solid Oxide Fuel Cells (SOFCs) are promising candidates to coach the energy transition. Yet, SOFC performance are markedly affected by fuel composition and operative parameters. In order to optimize SOFC operation and to provide a prompt regulation, reliable performance simulation tools are required. Given the high variability ascribed to the fuel in the wide range of SOFC applications and the high non-linearity of electrochemical systems, the implementation of artificial intelligence techniques, like Artificial Neural Networks (ANNs), is sound. In this paper, several network architectures based on a feedforward-backpropagation algorithm are proposed and trained on experimental data-set issued from tests on commercial NiYSZ/8YSZ/LSCF anode supported planar button cells. The best simulator obtained is a 3-hidden layer ANN (25/22/18 neurons per layer, hyperbolic tangent sigmoid as transfer function, obtained with a gradient descent with adaptive learning rate backpropagation). This shows high accuracy (RMS = 0.67% in the testing phase) and successful application in the forecast of SOFC polarization behaviour in two additional experiments (RMS in the order of 3% is scored, yet it is reduced to about 2% if only the typical operating current density range of real application is considered, from 300 to 500 mA·cm−2). Therefore, the neural tool is suitable for system simulation codes/software whether SOFC operating parameters agree with the input ranges (anode feeding composition 0–48%vol H2, 0–38%vol CO, 0–45%vol CH4, 9–32%vol CO2, 0–54%vol N2, specific equivalent hydrogen flow-rate per unit cell active area 10.8–23.6 mL·min−1·cm−2, current density 0–1300 mA·cm−2 and temperature 700–800 °C).