Adaptive Neuro-fuzzy Inference System Prediction Research Articles

Modeling of rheological behavior of honey using genetic algorithm–artificial neural network and adaptive neuro-fuzzy inference system

Abstract Knowledge of rheological properties of honey is of great interest to honey handlers, processors and keepers. In this study, genetic algorithm–artificial neural network (GA–ANN) and adaptive neuro-fuzzy inference system (ANFIS) models were used to predict the viscosity of four types of honey, two poly floral (Mountain, Forest) and two monofloral (Sunflower, Ivy). The GA–ANN and ANFIS were fed with 3 inputs of water content (15.25–19.92%), temperature (10–30 °C) and shear rate (1–42 s−1). The developed GA–ANN, which included 11 hidden neurons, could predict honey viscosity with correlation coefficient of 0.997. The overall agreement between ANFIS predictions and experimental data was also very good (r=0.999). Sensitivity analysis results showed that temperature was the most sensitive factor for prediction of honey viscosity. Both GA–ANN and ANFIS models predictions agreed well with testing data sets and could be useful for understanding and controlling factors affecting viscosity of honey.

ESTIMATION OF TOTAL DISSOLVED SALTS OF IRRIGATION WATER USING ADAPTIVE NEURO FUZZY INFERENCE SYSTEM

There are numerical indicators indicating the suitability of water for agriculture irrigation. These indicators focus on total dissolved salts (TDS) as the crops can be grown are dependent on such indicator. However, it has impact on crop productivity. The laboratory analysis is usually used in TDS determenation or TDS can be determined through empirical equations by measuring the electrical conductivity of the irrigation water. On the other hand, the empirical equations did not reflect the actual situation of the used water. So, in this study, an Adaptive Neuro Fuzzy Inference System (ANFIS) was employed to estimate the concentration of total dissolved salts of irrigation water. Howevere, the water samples were collected from wells located in Huraimla Governorate, Saudi Arabia. ANFIS Sugeno model was used to formulate the system. The inputs to the ANFIS were pH, calcium, magnesium, nitrate and iron concentrations and the output was TDS concentration. The possibility of the developed system to estimate the concentration of TDS was proven through comparing with results from multiple linear regression model. The mean relative error between actual TDS and estimated TDS was 2.972% for testing data set when using ANFIS in prediction. While, by using multiple linear regression model in prediction of TDS, the mean relative error was -15.782% for testing data set between actual TDS and estimated TDS. Thus it is possible to use ANFIS to estimate the concentration of total dissolved salts and it can be used as a management tool for water irrigation purposes.

Comparative Modeling and Analysis of the Mass Transfer Coefficient in a Turbulent Bed Contactor using Artificial Neural Network and Adaptive Neuro-Fuzzy Inference Systems

This study deals with predicting the gas film volumetric mass transfer coefficient (Koga) in a turbulent bed contactor (TBC), using both artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) techniques. The networks have been trained and evaluated with the experimental data available in the literature. Input variables to the networks are process variables such as gas and liquid phase concentration, gas and liquid superficial velocities and also specific area of packings. The results obtained the ability of developed ANN and ANFIS for prediction of Koga. Although it was observed that both ANN and ANFIS models provided a good statistical prediction in terms of coefficient of determination (R2), mean relative error (MRE) and root mean square error (RMSE), the accuracies of ANFIS predictions were better than those of ANN predictions.

Genetic algorithm-artificial neural network and adaptive neuro-fuzzy inference system modeling of antibacterial activity of annatto dye on Salmonella enteritidis

Annatto is commonly used as a coloring agent in the food industry and has antimicrobial and antioxidant properties. In this study, genetic algorithm-artificial neural network (GA-ANN) and adaptive neuro-fuzzy inference system (ANFIS) models were used to predict the effect of annatto dye on Salmonella enteritidis in mayonnaise. The GA-ANN and ANFIS were fed with 3 inputs of annatto dye concentration (0, 0.1, 0.2 and 0.4%), storage temperature (4 and 25°C) and storage time (1–20 days) for prediction of S. enteritidis population. Both models were trained with experimental data. The results showed that the annatto dye was able to reduce of S. enteritidis and its effect was stronger at 25°C than 4°C. The developed GA-ANN, which included 8 hidden neurons, could predict S. enteritidis population with correlation coefficient of 0.999. The overall agreement between ANFIS predictions and experimental data was also very good (r=0.998). Sensitivity analysis results showed that storage temperature was the most sensitive factor for prediction of S. enteritidis population.

Quantitative structure–activity relationship analysis of human neutrophil elastase inhibitors using shuffling classification and regression trees and adaptive neuro-fuzzy inference systems

The purpose of this study was to develop quantitative structure–activity relationship models for N-benzoylindazole derivatives as inhibitors of human neutrophil elastase. These models were developed with the aid of classification and regression trees (CART) and an adaptive neuro-fuzzy inference system (ANFIS) combined with a shuffling cross-validation technique using interpretable descriptors. More than one hundred meaningful descriptors, representing various structural characteristics for all 51 N-benzoylindazole derivatives in the data set, were calculated and used as the original variables for shuffling CART modelling. Five descriptors of average Wiener index, Kier benzene-likeliness index, subpolarity parameter, average shape profile index of order 2 and folding degree index selected by the shuffling CART technique have been used as inputs of the ANFIS for prediction of inhibition behaviour of N-benzoylindazole derivatives. The results of the developed shuffling CART-ANFIS model compared to other techniques, such as genetic algorithm (GA)-partial least square (PLS)-ANFIS and stepwise multiple linear regression (MLR)-ANFIS, are promising and descriptive. The satisfactory results ( = 0.845, = 0.861, = 0.829, RMSELOO = 0.305 and RMSEL25%O = 0.336) demonstrate that shuffling CART-ANFIS models present the relationship between human neutrophil elastase inhibitor activity and molecular descriptors, and they yield predictions in excellent agreement with the experimental values.

Investigation on the Effect of Data Imbalance on Prediction of Liquefaction

AbstractData imbalance causes learning bias in class identification techniques. A major cause for limited success in the prediction of liquefaction potential by various pattern recognition techniques is because of a liquefaction to nonliquefaction data class imbalance. It is suggested to use a support vector data description (SVDD) strategy to compensate the minority data. SVDD is used to generate virtual data points for the minority class bearing the same characteristics as the nonvirtual samples. Then an adaptive neuro-fuzzy inference system (ANFIS) classifier is employed to determine the liquefaction threshold. The ANFIS predictions are then examined by evaluating the coefficient of determination (COD) and comparing it with the Bayesian updating method. It is shown that for the liquefied data the approach is as efficient as the Bayesian method, but great improvement in the recognition rates of the nonliquefied data have been achieved.

A Novel Data Preprocessing Method for the Modeling and Prediction of Freeze-Drying Behavior of Apples: Multiple Output–Dependent Data Scaling (MODDS)

In the present study, the freeze drying behavior of apples have been modeled and predicted. Because freeze-drying is a very expensive and complex process, modeling of the freeze-drying process is a challenging task. In this study, a novel data scaling method called multiple output–dependent data scaling (MODDS) has been proposed and combined with an adaptive neuro-fuzzy inference system (ANFIS) to forecast the moisture content (MC), moisture ratio (MR), and drying rate (DR) values, which are outputs of freeze-drying behavior of apples. The input parameters of the freeze drying system are the sample thicknesses, drying time, pressure, relative humidity, chamber temperature, and sample temperature. Using the input parameters, the outputs of the freeze-drying process of apples were predicted using a hybrid system based on MODDS and ANFIS. In the first stage, only input parameters were scaled using MODDS. In the second stage, the outputs of freeze drying of apples were predicted with the scaled input parameters using ANFIS algorithm. Ninety-two samples were included in the data set, including 10-, 7-, and 5-mm samples. In order to evaluate the performance of the proposed model, the mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), coefficient of determination (R 2), index of agreement (IA), and mean absolute percentage error (MAPE) were used. Though MSE values of 2.48, 0.035, and 0.011 and IA values of 0.887, 0.887, and 0.466 were obtained for MC, MR, and DR, respectively, using the ANFIS prediction algorithm the hybrid MODDS-ANFIS model achieved MSE values of 0.003, 0.00005, and 0.00007 and IA values of 0.999, 0.999, and 0.993 for the prediction of MC, MR, and DR, respectively. The results obtained demonstrate that the proposed hybrid system is a robust and efficient method for the modeling and prediction of freeze-drying behavior of apples.

Membrane permeate flux and rejection factor prediction using intelligent systems

Backpropagation artificial neural network (BPNN), radial basis function (RBF) and adaptive neuro-fuzzy inference system (ANFIS) were utilized to predict starch removal performance from starchy wastewater using a hydrophilic polyethersulfone membrane with 0.65 μm pore size in a plate and frame homemade membrane module. Our study focuses on evaluation of membrane performance by optimum condition determination of operative parameters which affect the COD removal percentage and permeate flux. In this experiment, a four input vector was surveyed, including flow and temperature of feed, pH and concentration of permeate. In BPNN the number of neurons in the hidden layers needs to be chosen carefully to obtain a reliable network while choosing this structure is very time consuming. The best BPNN performance was obtained with 4 hidden layers for permeation and rejection factor prediction for BPNN. ANFIS and RBF simulations have also been used for comparison with BPNN. The results show a good agreement however the ANFIS prediction was better than two other simulation methods. In the basis of comparison between obtained results in this research, it may be an appropriate interpretation that for those chemical processes with performance which relied upon different variables, good performance prediction will be achieved by ANFIS systems.