regression (0.998742). It should also be observed that reliability prediction by TTF time series equally gave a considerably high value for the coefficient of determination of the linear regression line between the predicted values and the actual values (0.9824). Yet ANN model gave a better value. The closeness in the results from the two methods is not unconnected with the size of the data used for the work. ArtificialNeuralNetwork models are designed to handle stochastic and complex data analysis problems far beyond the size of example data used in this presentation. When the problem involves few data sets, statistical and probabilistic analysis tools can attempt to rival solutions offered by Artificial intelligence techniques. However, the ANN approach presented in this report still exhibited superior prediction capability over the TTF time series non- parametric analytic approach. This comparison can be done for large amount of data and other artificial intelligence techniques incorporated to fully elicit optimal reliability prediction methodology for deteriorating systems.
Glass fiber reinforced epoxy (GFRE) composite materials are prone to suffer from water absorption due to their heterogeneous structure. The main proc- ess governing water absorption is diffusion of water molecules through the epoxy matrix. However, hydrolytic degradation may also take place during components service life specially due high temperatures. In order to mitigate the effects of the water diffusive processes in the deterioration of in-service behavior of epoxy matrix composites, the use of chemically modified nano- clays as an additive has been proposed and studied in previous works [1]. In this work, an ArtificialNeuralNetwork (ANN) model was developed for bet- ter understanding and predicting the influence of modified and unmodified bentonite addition on the water absorption behavior of epoxy-anhydride sys- tems. An excellent correlation between model and experimental data was found. The ANN model allowed the identification of critical points like the precise temperature at which a particular system’s water uptake goes beyond a predefined threshold, or which system will resist an immersion longer than a particular time.
The advancement of modern technology and the speedy growth of human population has caused rapid expansion of energy consumptions. The necessity of efficient energy management and forecasting energy consumption know no bound. Developing large power system forecasting method using machine learning methods such as ArtificialNeural Networks (ANN) is a prospective approach for such purpose. In recent years, load forecasting has become one of the major areas of research in ArtificialNeuralNetwork. This paper presents a model of time-series based short-term load forecasting for the dataset collected from Regional Power Control Center of a Saudi Electricity Company. Due to the potential of the architecture to take the advantages of both time series and regression methods, ArtificialNeuralNetwork performs better than other learning methods. The proposed architecture is explored by the clustering of datasets based on k-means clustering approaches and hence proved that it indeed works.
The kinematic viscosity prediction model was developed for biodiesel using an ANN which represents a mathematical relationship between input and output parameters of a system as such a black box model. The selection of input parameters which contribute to the output is therefore a crucial task. It is also desirable to minimize the number of input parameters for an ANN system in order to reduce the computational time. In general, the best input parameters are being selected based on an understanding of the physics of a problem. Published literature suggested that the kinematic viscosity of biodiesel is a function of chemical composition- the fatty acid profile of methyl ester and impurities. In the present study, 27 variables are used as input parameters in developing the ANN for viscosity prediction. These parameters include: mass percentage of 18 fatty acid methyl ester that is commonly found in the biodiesel and 9 parameters listed in Table 2. Among 352 data sets that have been collected from literature, 327 sets are used in the ANN model training process whereas other 25 data sets are randomly selected for simulation.
In the last decade, several works introduced the use of artificialneuralnetwork (ANN) as a tool for automated sleep scoring. Most of the system used spectral in- formation of the signal using Fourier transformation [2]. Computerized EEG and other electrophysiological parameters monitoring reduces the problem of huge data handling. Computerization has led to more sophisticated use of EEG, even in effective disorders, where percep- tual processes are significantly distorted [3-5]. Fourier transformation, as a conventional method, has been widely used for the standard quantitative analysis of the spectral decomposition and the clinical application of EEG signals [6]. The ANN programs were developed for the analysis of most of the works that rely on spectral analysis and power spectrum method to evaluate elec- trographic data. In an attempt to classify sleep-wake stages determined, power of FFT or power spectrum band were used for better performance of the system [3,7]. The numerical values of the power of different frequency bands were used as inputs to ANN. As multi- layer perceptron neuralnetwork (MLPNN) undergoes some limitations, the performance of SOFM has also been tested to solve the problem at hand. In the present study, an effort has been made to exploit the inherent qualities of SOFM. The results obtained from computer simulations have been found to be very encouraging.
In Maharashtra state 27 districts are chosen and in that, publicly available records were taken, these records include parameters like minimum temperature, average temperature, maximum temperature, area, and production. For this dataset Multilayer Perceptron, NeuralNetwork is applied for processing with the help of WEKA tool. But this work suggests that in the ArtificialNeuralNetwork based model the prediction capabilities can be improved by considering additional parameters [7]. Amir Haghverdia, Robert A. Washington-Allenb, Brian G. Leibc predicted the cotton lint yield using remote sensing technology. The satellite remote sensing technology is primarily used for assessment and monitoring of the agricultural land in order to determine the area, amount and type of crop production. Deep Learning can be applied and used for this type of problems. In this ArtificialNeuralNetwork (ANN) approach is used to generate the models related to individual Crop Indices (CI) and CI phenology to map and predict the yield of cotton lint in two growing seasons.
Abstract: Partial discharge (PD) can provide a useful forewarning of asset failure in electricity substations. A significant proportion of assets are susceptible to PD due to incipient weakness in their dielectrics. This paper examines a low cost approach for uninterrupted monitoring of PD using a network of inexpensive radio sensors to sample the spatial patterns of PD received signal strength. Machine learning techniques are proposed for localisation of PD sources. Specifically, two models based on Support Vector Machines (SVMs) are developed: Support Vector Regression (SVR) and Least-Squares Support Vector Regression (LSSVR). These models construct an explicit regression surface in a high dimensional feature space for function estimation. Their performance is compared to that of artificialneuralnetwork (ANN) models. The results show that both SVR and LSSVR methods are superior to ANNs in accuracy. LSSVR approach is particularly recommended as practical alternative for PD source localisation due to it low complexity.
This paper analyses the performance of Artificialneuralnetwork (ANN) and Support vector machines (SVM) in terms of overall accuracy in detecting nonTor traffic in a Tor network traffic dataset data from the University of New Brunswick (UNB), Canadian Institute for cyber security (CIC) using an anomaly based approach with all features in the dataset. As part of the work, the results are compared with the results of [1] being the only study published to the best of our knowledge using the UNB- CIC Tor Network Traffic dataset. In the proposed approach 10 features are selected out of the 28 features of the dataset using Correlation based feature selection (CFS) for training and testing the detection algorithms.
portable water demand, which requires exploration of raw water sources, developing treatment systems. The main function of water treatment plants is to protect human health and the environment from excessive overloading of various pollutants. Coagulation is an essential part of drinking water treatment process allowing the removal of suspended colloidal particles in water. The chemical coagulation is the process of destabilizing the colloidal particles suspended in raw water by the addition of coagulants. The coagulant dosage rate is nonlinearly related to raw water characteristics such as turbidity, pH, and alkalinity. The most common coagulants used are Aluminum Sulfate (Alum). Optimizing coagulation for selecting the best coagulation condition and coagulation dosage is important. The main test used to determine the optimum coagulation conditions is the jar test, which requires a long experimental time. Modeling can be used to overcome this limitation. In this study, a model for approximation of coagulant dosage rate in water treatment plant in Punalur, Kollam has been developed using the ArtificialNeuralNetwork (ANN). The ANN is known as an excellent estimator of nonlinear relationships between accumulated input and output numerical data. Using this nature of the ANN, the optimal coagulant dosing rate can be predicted from the operating data with accuracy and in time. Optimization of the Coagulation process using ANN can reduce the amount of chemicals being used, testing time and consequently reduced the operational cost. Document
portable water demand, which requires exploration of raw water sources, developing treatment systems. The main function of water treatment plants is to protect human health and the environment from excessive overloading of various pollutants. Coagulation is an essential part of drinking water treatment process allowing the removal of suspended colloidal particles in water. The chemical coagulation is the process of destabilizing the colloidal particles suspended in raw water by the addition of coagulants. The coagulant dosage rate is nonlinearly related to raw water characteristics such as turbidity, pH, and alkalinity. The most common coagulants used are Aluminum Sulfate (Alum). Optimizing coagulation for selecting the best coagulation condition and coagulation dosage is important. The main test used to determine the optimum coagulation conditions is the jar test, which requires a long experimental time. Modeling can be used to overcome this limitation. In this study, a model for approximation of coagulant dosage rate in water treatment plant in Punalur, Kollam has been developed using the ArtificialNeuralNetwork (ANN). The ANN is known as an excellent estimator of nonlinear relationships between accumulated input and output numerical data. Using this nature of the ANN, the optimal coagulant dosing rate can be predicted from the operating data with accuracy and in time. Optimization of the Coagulation process using ANN can reduce the amount of chemicals being used, testing time and consequently reduced the operational cost. Document
In this work, the product advancement stage is to execute the framework that can guarantee the exhibition, ideal and adaptable endeavor asset arranging framework specifically Hybrid Analytic Hierarchy Process-ArtificialNeuralNetwork (HAHP-ANN) strategy. In the proposed technique, the mix approach with the association of the decision-maker at the pecking order of the procedure is to get the best enhancement. And furthermore, to improve the task customization, an incorporation of closest neighbor classifier and mix SVM for learning necessities is utilized. The general assessment of the examination strategy is led in the java simulation environment which it is demonstrated that the proposed research system prompts give the ideal result than the current research approaches.
Identification results of ANN model Artificialneuralnetwork (ANN) is a non-linear pattern recognition method. Many parameters exert to some extent certain influence on the perform- ance of ANN models. These parameters include the number of neurons in the middle layer, scale functions, learning rate factor, momentum factors, and initial weights (Blanco et al. 1999; Mouwen et al. 2006). In this work, the most classical Back Propagation ArtificialNeuralNetwork (BP-ANN) with 3 layers construction was used to construct the identification model. These parameters of the BP-ANN model were optimised by cross validation as follows: the number of neurons in the hidden layer was set to 8, the learning rate factor and momentum factor were set to 0.1 each, the initial weight was set to 0.3, and the scale function was set as ‘tanh’ function.
ArtificialNeuralNetwork (ANN) was developed to predict the Open Cicuit Potential (OCP) values and Nyquist plot for bare and bioploymer coated cp-Titanium substrate. The experimental data obtained was utilised for ANN training. Two input paramters, i.e., substrate condition (coated or uncoated) and time period were considered to predict the OCP values. Back propogation Levenberg-Marquardt training algorithm was utilised in order to train ANN and to fit the model. For Nyquist plot, the network was trained to predict the imaginary impedance based on real impedance as a function of immersion periods using the Back Propagation Bayesian
The flow behaviour of material during hot forming process is usually complicated. The hardening and softening mechanisms both mainly affect the strain rate and temperature [1–3]. The understanding of metallic alloys deformation behaviour at elevated temperatures helps to provide information about the metal forming processes. Three main categories of models are utilised to describe the stress flow behavior of metallic alloys: (1) physical based; (2) phenomenological and (3) artificialneuralnetwork constitutive models [4–8]. Phenomenological constitutive models are usually used in the simulation of hot forming processes due to their practicability and accuracy. For titanium alloys, the more significant part of literature on the constitutive modelling pays particular attention to α + β type alloys, especially Ti-6Al-4V titanium alloy [8–11]. The Arrhenius-type constitutive equation (ACE), where the flow
Intrusion detection has attracted a considerable interest from researchers and industry. After many years of research the community still faces the problem of building reliable and efficient intrusion detection systems (IDS) capable of handling large quantities of data with changing patterns in real time situations. The Tor network is popular in providing privacy and security to end user by anonymizing the identity of internet users connecting through a series of tunnels and nodes. This work identifies two problems; classification of Tor traffic and nonTor traffic to expose the activities within Tor traffic that minimizes the protection of users in using the UNB-CIC Tor Network Traffic dataset and classification of the Tor traffic flow in the network. This paper proposes a hybrid classifier; ArtificialNeuralNetwork in conjunction with Correlation feature selection algorithm for dimensionality reduction and improved classification performance. The reliability and efficiency of the propose hybrid classifier is compared with Support Vector Machine and naïve Bayes classifiers in detecting nonTor traffic in UNB-CIC Tor Network Traffic dataset. Experimental results show the hybrid classifier, ANN-CFS proved a better classifier in detecting nonTor traffic and classifying the Tor traffic flow in UNB-CIC Tor Network Traffic dataset.
Heart is the next major organ comparing to brain which has more priority in Human body. It pumps the blood and supplies to all organs of the whole body. Prediction of occurrences of heart diseases in medical field is significant work. Data analytics is useful for prediction from more information and it helps medical centre to predict of various disease. Huge amount of patient related data is maintained on monthly basis. The stored data can be useful for source of predicting the occurrence of future disease. Some of the data mining and machine learning techniques are used to predict the heart disease, such as ArtificialNeuralNetwork (ANN), Decision tree, Fuzzy Logic, K-Nearest Neighbour(KNN), Naïve Bayes and Support Vector Machine (SVM). This paper provides an insight of the existing algorithm and it gives an overall summary of the existing work.
Abstract: Rainfall is considered as one of the major components of the weather forecasting. In the current world climate change, the accuracy of rainfall forecasting model is very important factor. Rainfall affects the drought and flood situation. India is an agricultural country. Rainfall are also affects the area of agriculture. We have considered the monthly rainfall data of East Madhya Pradesh, India from 1901-2017. Afterwards, in this paper, to evaluate an actual prediction of rainfall forecasting, were used an ArtificialNeuralNetwork (ANN) by the Cascade-forward Back propagation NeuralNetwork (CFBPNN) technique. In this study, to trained the training data of the rainfall information using 2 hidden layers of CFBPNN technique, with three different epochs: [2-50-10-1] with epoch fix to 500, [2-50-20-1] with epochs fix to 1000 and 1500. To measure the performance of the developed architecture, the mean square error (MSE) algorithm is employed using CFBPNN. The experimental results showed that [2-50-10-1] architecture with epoch fix to 500 and learning rate 0.1, produced a good performance result with the value of MSE was 0.0063408. Eventually, CFBPNN algorithm has provided a best accuracy model to predict monthly rainfall in East Madhya Pradesh, India.
Abstract — This paper describes the development of an adaptive control mechanism for FES-assisted indoor rowing exercise (FES-rowing). The FES-rowing is intro- duced as a total body exercise for rehabilitation of function of lower body through the application of functional elec- trical stimulation (FES). A model of the rowing ergometer with humanoid is developed using the visual Nastran soft- ware environment (vN4D). A fuzzy logic control (FLC) scheme is designed in Matlab/Simulink and adapted online by pre-training artificialneuralnetwork (ANN) to regulate the muscle stimulation pulse width required to drive FES-rowing. The ANN is used as an adaptation to the system that is required to account for muscle fatigue. The results signify that the adaptive control scheme is able to achieve and maintain better tracking performance. This study indicates that the adaptive control developed may provide an effective mechanism for automatically regulat- ing the stimulation pulse width for FES-rowing to over- come muscle fatigue.
Abstract: Blood is the very important biological fluid present in human beings, which has various functions. Leukaemia is a blood disease which affects the normal functioning of the blood. The number of cases of leukaemia reported every year in India is approximately close to 1 million. Detection and classification of Leukaemia into its subtypes manually by humans is difficult and time consuming. Hence an automatic system to identify and classify Leukaemia may overcome this drawback. Technique proposed in this paper aims at quick detection of leukaemia images, and classifies them into their respective subtype. This is achieved using image processing with the help of MATLAB software. Images used are microscopic blood smear images. Using K- means clustering algorithm images are segmented and a ArtificialNeuralNetwork (ANN) is designed based on the features extracted. This neuralnetwork is trained to classify the images into their respective type.
The prediction accuracy may be altered by the presence of irrelevant or redundant attributes. We will perform two types of feature selection in order to improve the classification accuracy and the total computation time. The first one, is done before the clustering phase, in which we will use the Fast Correlation-Based Filter, described here [13] in order to eliminate irrelevant features, the ones taking only one single value for example, carrying no information about input vectors.