Sharif Digital Repository

The application of artificial neural network to compressor performance map prediction is investigated. Different types of artificial neural network such as multilayer perceptron network, radial basis function network, general regression neural network, and a rotated general regression neural network proposed by the authors are considered. Two different models are utilized in simulating the performance map. The results indicate that while the rotated general regression neural network has the least mean error and best agreement to the experimental data, it is however limited to curve fitting application. On the other hand, if one considers a tool for curve fitting as well as for interpolation... 

The usual key assumptions in designing quality control charts are the normality and independency of serial samples. While the normality assumption holds in most cases, in many continuous-flow processes such as the chemical processes, serial samples have some degrees of autocorrelation associated with them. Ignoring the autocorrelation structure in constructing control charts, results in decreasing the in-control run length, and so increasing the false alarms. Moreover, when the object is to detect small shifts in the mean vector of a process, the performance of Cumulative Sum (CUSUM) control charts is dramatically better than Schewhart control charts. One of the methods, which have been... 

A set of conventional feedforward multilayer neural networks have been proposed to predict basic properties (e.g., critical temperature (T c), critical pressure (Pc), critical volume (V c), acentric factor (ω) and molecular weight (MW)) of pure compounds and petroleum fractions based on their normal boiling point (T b) and liquid density at 293 K. The accuracy of the method is evaluated by its application for basic property estimation of various components not used in the development of the method. Furthermore, the performance of the method is compared against the performance of the other alternatives reported as the most accurate and general methods for basic property prediction. Results of... 

In this paper we will address two major problems in mammogram analysis for breast cancer in MR-images. The first is classification between normal and abnormal cases and then, discrimination between benign and malignant in cancerous cases. Our proposed method extracts textural and statistical descriptive features that are fed to a learning engine based on the use of Support Vector Machine learning framework to categorize them. The obtained results show excellent accuracy in both classification problems, that proves the appropriate combination of our features and selecting powerful classifier i.e. Support Vector Machine leads us to a brilliant outcome  

Neural networks have been used for process fault diagnosis. In this work, the cluster analysis is used to design a structurally optimized Probabilistic Neural Network. This network is called the Clustered-Based Design Probabilistic Neural Network (CBDPNN). The CBDPNN is capable of diagnosing the faults of nonlinear processes operating over several regions. The performance and training status of the proposed CBDPNN is compared to a conventional Multi-Layer Perceptron (MLP) that is trained on the whole operating region. Simulation results indicate that both schemes have the same performance, but, the training of CBDPNN is much easier than the conventional MLP, although it has about 50% more... 

This paper presents a scalp electroencephalogram (EEG) rhythmic pattern detection scheme based on neural networks. Rhythmic discharges detection is applicable to the majority of seizures seen in newborns, and is listed as detecting 90% of all the seizures. In this approach some features based on various methods are extracted and compared by a modified multilayer neural network in order to find rhythmic discharges. Statistical performance comparison with seizure detection schemes of Gotman et al. and Liu et al. is performed. © 2006 IEEE  

The purpose of this investigation is to develop a model for prediction of topologically closed-packed (TCP) phases formation in superalloys. In this study, artificial neural networks (ANN), using several different network architectures, were used to investigate the complex relationships between TCP phases and chemical composition of superalloys. In order to develop an optimum ANN structure, more than 200 experimental data were used to train and test the neural network. The results of this investigation shows that a multilayer perceptron (MLP) form of the neural networks with one hidden layer and 10 nodes in the hidden layer has the lowest mean absolute error (MAE) and can be accurately used... 

This paper concerns the design of a neural state observer for nonlinear dynamic systems with noisy measurement channels and in the presence of small model errors. The proposed observer consists of three feedforward neural parts, two of which are MLP universal approximators, which are being trained off-line and the last one being a Linearly Parameterized Neural Network (LPNN), which is being updated on-line. The off-line trained parts are able to generate state estimations instantly and almost accurately, if there are not catastrophic errors in the mathematical model used. The contribution of the on-line adapting part is to compensate the remainder estimation error due to uncertain parameters... 

In this paper, we propose a novel OCR system which can recognize and calculate handwritten Persian arithmetic expressions without using a keyboard or a memory to store the intermediate results. Our system is composed of two major phases: character recognition and calculation. The recognition phase is based on a new approach for feature extraction followed by a Fuzzy Support Vector Machines (FSVMs) as the classifier. In calculation phase a simple algorithm is used for calculating the recognized arithmetic expressions. The performance of the system was evaluated on a database consisting of 3400 digits and symbols written by 20 different people. 92 percent accuracy in recognition proves the... 

The belief propagation algorithm is a state of the art decoding technique for a variety of linear codes such as LDPC codes. The iterative structure of this algorithm is reminiscent of a neural network with multiple layers. Indeed, this similarity has been recently exploited to improve the decoding performance by tuning the weights of the equivalent neural network. In this paper, we introduce a new network architecture by increasing the number of variable-node layers, while keeping the check-node layers unchanged. The changes are applied in a manner that the decoding performance of the network becomes independent of the transmitted codeword; hence, a training stage with only the all-zero... 

Internet-of-Things (IoT) devices and applications are being deployed in our homes and workplaces. These devices often rely on continuous data collection to feed machine learning models. However, this approach introduces several privacy and efficiency challenges, as the service operator can perform unwanted inferences on the available data. Recently, advances in edge processing have paved the way for more efficient, and private, data processing at the source for simple tasks and lighter models, though they remain a challenge for larger and more complicated models. In this article, we present a hybrid approach for breaking down large, complex deep neural networks for cooperative, and... 

As recent neural networks are being improved to be more accurate, their model's size is exponentially growing. Thus, a huge number of parameters requires to be loaded and stored from/in memory hierarchy and computed in processors to perform training or inference phase of neural network processing. Increasing the number of parameters causes a big challenge for real-time deployment since the memory bandwidth improvement's trend cannot keep up with models' complexity growing trend. Although some operations in neural networks processing are computational intensive such as convolutional layer computing, computing dense layers face with memory bandwidth bottleneck. To address the issue, the paper... 

An artificial neural network model was developed for modeling of the effects of mechanical alloying parameters including milling time, milling speed and ball to powder weight ratio on the characteristics of Al-8 vol%SiC nanocomposite powders. The crystallite size and lattice strain of the aluminum matrix were considered for modeling. This nanostructured nanocomposite powder was synthesized by utilizing planetary high energy ball mill and the required data for training were collected from the experimental results. The characteristics of the particles were determined by X-ray diffraction, scanning and transmission electron microscopy. Two types of neural network architecture, i.e. multi-layer... 

The phenomenon of the hydraulic jump is so complex that despite considerable laboratory and prototype studies, estimation of its main characteristics in a generalized and accurate form is still difficult. The Artificial Neural Network (ANN) approach aims at limiting the needs for costly and time-consuming experiments. In this study, two ANN models, multi-layer perceptron using back propagation algorithm (MLP/BP) and radial basis function using orthogonal least-squares algorithm (RBF/OLS), were used to predict the roller length, sequent depth, and the relative energy loss of the B-jump. Based on a pre-specified range of jump parameters, the input vectors include: upstream bed slope (tan θ),... 

Modular Neural Networks have had significant success in a wide range of applications because of their superiority over single non-modular ones in terms of proper data representation, feasibility of hardware implementation and faster learning. This paper presents a constructive multilayer neural network (CMNN) in conjunction with a Hopfield model using a new cost function to simulate the behavior of superconductive fault current limiters (SFCLs). The results show that the proposed approach can efficiently simulate the behavior of SFCLs. ©2006 IEEE  

Evaporative condensers are regarded as highly-efficient and eco-friendly heat exchangers in refrigeration systems. Data-driven methods can play a key role in performance prediction of evaporative condensers, conducted without the complexity of theoretical analysis. In this study, four machine learning models including multi-layer perceptron artificial neural network (ANNMLP), support vector regression (SVR), decision tree (DT), and random forest (RF) models have been employed to predict heat transfer rate and overall heat transfer coefficient of a small-scale evaporative condenser functioning under a wide range of working conditions. A set of experimental tests were conducted, where inlet...