The vast weather changes effect on human activities. Dealing with weather data manually is very difficult job and time consuming operation. The process of data entry requires a precise method suits different weather parameters. Artificial intelligent [AI] especially, hybrid systems improve the performance of either pure neural network based or pure fuzzy logic based forecasting. In this study, a Neuro-fuzzy approach will be proposed to predict weather in Sadat region, western desert, Egypt. A combination of monthly mean meteorology measurements for temperature, relative humidity, wind speed, and rainfall will be used during the period [2008-2017]. Many methods were applied over the years for weather prediction such as classical and intelligent techniques. The proposed model uses a Neuro-fuzzymodel at different types of fuzzy member ship functions. The flexibility of the proposed model increase the prediction accuracy. The effectiveness of the proposed model is demonstrated at different operating conditions. The classification of data is divided into 12 sets; each set consists of 4 mean values of observations. A transposing process applied on these sets for training and testing at different number of rules 10, 11, 15, 20, 25, 30, 35, and 40. Eight choices for membership functions "triangular" and another for "Gaussian" performed. The accuracy of the output forecasting measured using MAPE and MAE. A comparison applied among different cases obtained from Neuro-fuzzymodel and observed meteorological data for year 2017. The results show that the performance of the Neuro- fuzzymodel at TCWB is better than TLWB. Also, the model at GCWB and GCWN are better than GCCB and GCCN. The results show that Neuro-fuzzymodel seemed to be promising method for weather prediction.
Abstract— This paper presents a novel approach for segmentation based on Neuro-Fuzzymodel using decision making. We know that segmentation is done based on some feature values of images. These features work as parameters. There are many segmentation techniques available presently based on different approaches. Many of them require parameter selection which is done manually by observation. This can be performed on those data which are having easily differentiable values. But images such as skin lesion images have very marginal or not-differentiable data of lesion and skin which cannot be easily analysed. This makes parameter selection and assigning parameter value task very difficult. Hence, to solve this problem, we are presenting a novel approach for segmentation problem that uses decision making. We have evaluated this approach by applying it on different dermatological images containing skin lesion which results in good quality segmentation of skin lesion.
Abstract — It is difficult to identify the abnormalities in brain specially in case of Magnetic Resonance Image brain image processing. This paper presents a hybrid technique for the classification of MRI human brain images. The proposed hybrid technique consists of three stages namely feature extraction, feature reduction and classification. The feature extraction and reduction is done by Principal Component Analysis and the classification is done by a hybrid Neuro- fuzzy classifier (ANFIS). ANFIS classifier combines the merits of both the neuro classifier and the fuzzy classifier and overcomes the demerits of both the classifiers. Artificial neural networks employed for brain image classification are being computationally heavy and also do not guarantee high accuracy. The major drawback of ANN is that it requires a large training set to achieve high accuracy. On the other hand fuzzy logic technique is more accurate but it fully depends on expert knowledge, which may not always available. Fuzzy logic technique needs less convergence time but it depends on trial and error method in selecting either the fuzzy membership functions or the fuzzy rules. These problems are overcome by the hybrid model namely, neuro- fuzzymodel. This system removes essential requirements since it includes the advantages of both the ANN and the fuzzy logic systems. In this paper the classification of different brain images using Adaptive neuro-fuzzy inference systems (ANFIS technology) is done. Experimental results illustrate promising results in terms of classification accuracy and convergence rate.
The algorithm proposed by Castanho et al.  tunes the membership functions using a genetic algorithm, whereas we have used the Adaptive NeuroFuzzy Inference System (ANFIS) to optimise the membership functions. Furthermore, Castanho et al.’s  and our proposed system both aim to predict whether a patient has organ-confined disease (OCD, pathological stage pT2) or extra-prostatic disease (ED, pathological stage > pT2). Although both systems use pre-operative serum PSA, clinical stage, and primary and secondary Gleason grades of a biopsy to predict the pathological stage of prostate cancer, our system considers age as an additional input variable. Castanho et al. ’ s  genetic-fuzzy system achieved an Area Under the Curve of 0.824 which they compared against Partin probability tables which have been proposed by Makarov et al. , and which only achieved an Area Under the Curve of 0.693. Our proposed neuro-fuzzy approach achieved an Area under the curve of 0.812, and the AJCC nomogram achieved an Area Under the Curve of 0.582. These results approximate to those reported by Castanho et al. , and reveal a high degree of consistency among the two outcomes of the two studies, despite the fact that different datasets were used for each study. The nomograms used by Castanho et al., and the AJCC nomogram both use the TNM Classification of Malignant Tumors grading system . A major limitation of the AJCC nomogram is that the biopsy Gleason 7 values are not split into 3 + 4 = 7 vs. 4 + 3 = 7 which have drastically different clinical outcomes. The proposed neuro-fuzzymodel considers the Gleason Grades 3 + 4 and 4 + 3, and this was one of the reasons that it performed better than the AJCC nomogram.
In this work, comparison between experimental reading and Neuro-Fuzzymodel is carried out to define the optimal model for predicting the radiation levels of airborne radon and thoron in two Egyptian phosphate mines. Two cases for predicting radon and thoron levels are investigated in Safaga Omelhoytat and Safaga South mine. In the first case thoron and radon reading is taken at distances in series from the opening of the mine till the middle and the rest of the data are predicted. In the second case three random reading is taken at the opening, middle, end of the mine and the radioactivity measurements are predicted in the distances between the reading value. MAE and MAPE are calculated to assess and compare performance of the two cases using Neuro-Fuzzymodel. In Safaga Omelhoytat mine, the MAE and MAPE are 0.03 WL, 2.83% for radon and 0.0019 WL, 4.679% respectively, for thoron in the first case (series reading). While the MAE and MAPE for radon and thoron respectively are 0.0151WL, 1.448% and 7.16x10 -4 WL,
Blood pressure is the most often measured and intensively studied in medical and physiological practice because it is the valuable diagnostic aid to access the vascular condition of certain illnesses. As known that high blood pressure is the biggest known cause of disability and premature death through stroke, heart attack and heart disease. In this case, accurate blood pressure measurement and prediction need to be conducted. This project is mainly concerned on designing a model that can predict the Mean Arterial Pressure. By this project development, SPSS software used to decide the parameter that correlate with the blood pressure parameter based on Mean Arterial Pressure. Then utilizes the Neuro-Fuzzymodel using MATLAB as the mainframe. This project is based on certain age and a number of subjects patient.
Neural network has been found as an efficient method among many biometric techniques. It has the method of reducing the search space. Some of the images in the database are used for training and the rest are used for testing. As the number of images used for training increases the recognition efficiency is also increases. By using high speed systems with parallel processing capabilities this can be extended for a very large database. The main factor which affects the performance of the system is the time spent in training. But high speed or parallel processing systems can overcome this constraint as they can perform the training faster. As compared to NN approach, the NeuroFuzzymodel provides better results. The improved accuracy is due to its nearness to real world condition because of the use of probability factor in the assignment of desired output.
Failure trend in object-oriented programming courses is mostly on the increase side, student’s performance in other courses are most times better than in programming courses. One of the ways to improve the student performance is for the instructors to identify the group of students who might not perform well at the later stage of learning. From there the instructor can focus on the students in order to help them to improve their performance. Thus, in this case making the prediction of student performance a major step in identifying the potential students that needs further help such as extra classes or special tutorials and assignments. Therefore, the need for performance prediction in programming courses becomes imperative. The study will use neuro-fuzzymodel to evaluate the current performance of students and then predict the students’ performances in subsequent object oriented programming courses.
The aim of this work and its main contributions are in the development of a procedure and a model based on frequency domain analysis combined with the use of hybrid artificial intelligence techniques. The model estimates information from a high-range of signal frequencies with high levels of uncertainty in complex electro-mechanical processes. It is applied to estimate the micro-scale eccentricity of the spinning axis of a rotating device with ultra-precision performance requirements. This method can be effectively applied to reduce systemic error and decrease production time in manufacturing processes.
NeuroFuzzy designs focus on specific observations and enhance the effectiveness rating of decision making. Moreover, NeuroFuzzy design doesn't call for a priori knowledge of weights for inputs & outputs. Nevertheless, managerial judgment could be employed when it's desired . In 1965, Zadeh printed the very first paper on the novel method of characterizing nonprobabilistic uncertainties, which he named fuzzy sets, that has today turned into a fruitful place that contains different disciplines , like calculus of fuzzy if then rules, fuzzy graphs, fuzzy interpolation, fuzzy topology, fuzzy reasoning, fuzzy inferences methods, and fuzzy modeling. The apps, which are multi disciplinary in nature, includes automated command, decision-making, data classification, computer vision, database management, information retrieval, time-series prediction, signal processing, consumer electronics, etc. Certain fuzzy methods are universal feature approximations. To be able to determine a good fuzzy program for a certain problem, a principle along with membership functions base system. This is often accomplished by prior knowledge, by learning, and by a mix of both. In case a learning algorithm is used that utilizes community info and Causes neighborhood modifications in a fuzzy phone system. This approach is generally called Neuro-Fuzzy methods for identification of linear time invariant systems. The linear identification is grounded on measured input and production values of the method. Identification for nonlinear systems is based on measured input as well as output values, though it's harder . Authors have presented a process which uses neural networks for control as well as identification of nonlinear systems. For identification, the input as well as production values of the device are given right
An ANFIS is a multi-layered feed forward network, in which each layer performs a particular task. The layers are characterized by the fuzzy operations they perform. Fig- ure 2 describes the global architecture of this neural-fuzzy system. It shows a n-input, type-5 ANFIS. Three member- ship functions are associated with each input.
Abstract: Neuro-fuzzy techniques are finding a practical application in many fields such as in model identification and forecasting of linear and non-linear systems. This paper presents a neuro-fuzzymodel for forecasting the fruit production of some agriculture products (olives, lemons, oranges, cherries and pistachios). The model utilizes a time series of yearly data. The fruit forecasting is based on Adaptive Neural Fuzzy Inference System (ANFIS). ANFIS uses a combination of the least-squares method and the backprobagation gradient descent method to estimate the optimal food forecast parameters for each year. The results are compared to those of an Autoregressive (AR) model and an Autoregressive Moving Average model (ARMA). Keywords: Fruit forecasting, neuro-fuzzy, ANFIS, AR, ARMA, forecasting, fruit production
Abstract: Stock Market is considered as one of the fundamental building block of developed country so if number of inverstors increases then the economy of the country also increases and every investor invests to get good returns. But as stock market is uncertain and complicated the selection of good scripts are considered as one of the challenge in stock market field.This problem can be modelled using the Neuro-Fuzzy technique that can handle non-linearity and uncertainty in the stock market. Authors proposes a Neuro–Fuzzymodel using financial indicators which play a vital role in selection of scripts. In This model past quarter results of selected listed scripts of BSE India are considered for training and setting the parameters of Fuzzy Inference System (FIS) which could signal investors to have profitable script in their portfolio.
Prediction will provide a powerful tool for managers to be more successful in the long and short term planning for their organization. Prediction can be done in two ways: it can either be the result of deduction and analysis of an expert in a given field of knowledge, or the analysis and evaluation of raw data and statistics. In this study we consider prediction using time series data. Time series show different trends in different cases. If we want to divide this behavior into two general categories, we can say that data have either linear or non-linear trend. The purpose of this study is to analyze Auto-Regressive Integrated Moving Average model and artificial neural network model in fuzzy systems. Then with comparing these models we can conclude whether the classical Auto-Regressive Integrated Moving Average Model has the same prediction power as the neuro-fuzzymodel or not. Classical model of moving average or Box-Jenkins model, have conventionally been used with data having linear trend. But in the real world there are fewer cases where data having linear trend or static state in average and variance, so recently more accurate methods of modeling non-linear systems are invented.
Depression is a psychological disorder, which, if untreated, may deteriorate the quality of one’s life. Therefore, to tackle it, its early screening and accurate grading are much needed. The success of soft computing largely stands on its effective ways of handling uncertainty, which is often encountered in a clinical diagnosis. This paper proposes application of soft computing techniques to automate depression diagnosis. In order to achieve our goal, an intelligent Neuro-Fuzzymodel has been developed. It has been trained with a sample of real- world depression data. Experiments with test data reveal that the Mean Squared Error in prediction is nominal for most of the cases. Such a system could assist the doctors to take decisions in much needed situations.
The experiment is conducted with the 17 programs of Glace EMR Medical Billing. The analysis is done using FIS (fuzzy interference system) and the proposed NeuroFuzzymodel in MATLAB environment. The Fuzzy Inference Systems based model is presented with the parameter Normalized MTBF in Figure 4.7. The generated membership function based on employed fuzzy IF-THEN rules ranging from “very low” to “very high” are shown in Figure 4.8. The dataset will be given as training data to ANFIS systems and the
There are a large number of applications like medical imaging, video surveillance and remote sensing etc. that require images with both spatial and spectral resolution as well. In this paper, the potentials of image fusion using fuzzy and neurofuzzy approaches has been explored along with quality assessment evaluation measures. Fused images are primarily used to human observers for viewing or interpretation and to be further processed by a computer using different image processing techniques. All the results obtained and discussed by this method are same scene. The experimental results clearly show that the proposed image
ABSTRACT: Condition monitoring systems using vibration measurements and supervised classifiers can be used to auto- mate the diagnosis process of rotating machines. In this paper, we describe an automatic diagnosis system for detection and classification of defects in ball bearings using a time varying parametric spectrum estimation method for analyzing nonstationary vibration signals. The classification task is accomplished by an adaptive neural fuzzy inference system. The designed system was developed to be able to classify four types of preestablished defects in ball bearings operating under several shaft speeds and load conditions. The system was tested with experimental data collected from drive end ball bearing of an induction motor driven by mechanical system.
The neuro-fuzzy system uses the linguistic knowledge of fuzzy inference system and the learning capability of neural network. To describe the architecture a neurofuzzy system, consider Figure 1. For simplicity, we assume that our fuzzy system has two inputs and one output. Furthermore, we assume that the defuzzification of the variables is a linear combination of the first order of input variables (approach of Takagi - Sugeno).
ABSTRACT: This research was carried out to design a hardware and a software of the GPS based on GSM network in order to evaluate the implemented system experimentally. The proposed GPS/GSM based system model includes two parts which are the mobile unit and the control station. The mobile unit consists of Personal Computer PC-A connected to GSM modem- A and GPSreceiver through the computer’s USB ports, while the Control Station consists of another Personal Computer PC- B connected to GSM modem-B. This setup provides communication between PC-A and PC-B via SMS protocol. The system processes, interfaces, connections, data transmission and reception of data between the mobile unit and the control station has been carried out and tested successfully and accuracies less than 9m and 0.6m are obtained compared with the original coordinates using ordinary GPS and differential GPS respectively.