Many existing on-line testing systems cannot create adaptive examination papers that are fit for every student respectively. For traditional algorithms, the process of creating examination questions from question bank...
详细信息
ISBN:
(纸本)0972147950
Many existing on-line testing systems cannot create adaptive examination papers that are fit for every student respectively. For traditional algorithms, the process of creating examination questions from question bank is lack of high efficiency and occupies too much memory and CPU time. The speed of data transmission is very slow when the flow of information in network is very high. In order to solve these problems, a novel architecture of a personalized on-line testing system is proposed in this paper. Firstly, the mobile agent technology is used in this system architecture to mitigate the burden of server and the flow of information in network. The tasks in different phrases of an testing procedure are assigned to several agents: master agent, paper generation agent, mobile agent, interface agent, testing agent, evaluation agent and result agent respectively. These agents can collaborate with each other during the exam period, which arranges the testing procedure in an orderly way. Secondly, we use the back propagation algorithm to evaluate the student's teaming status roundly. Thirdly, we present an improved genetic algorithm. By means of these algorithms, the system can provide objective assessments and personalized examination papers for students.
Automatic feature extraction and classification algorithm of echo signal of ground penetrating radar is presented. Dyadic wavelet transform and the average energy of the wavelet coefficients are applied in this paper ...
详细信息
Automatic feature extraction and classification algorithm of echo signal of ground penetrating radar is presented. Dyadic wavelet transform and the average energy of the wavelet coefficients are applied in this paper to decompose and extract feature of the echo signal. Then, the extracted feature vector is fed up to a feed forward muhi layer perceptron classifier. Experimental results based on the measured GPR, echo signals obtained from the Mei shan railway are presented.
Artificial neural networks can be used as a fault diagnostic tool in chemical process industries. Connection strengths representing correlation between inputs (sensor measurements) and outputs (faults) are made to lea...
详细信息
Artificial neural networks can be used as a fault diagnostic tool in chemical process industries. Connection strengths representing correlation between inputs (sensor measurements) and outputs (faults) are made to learn by the network using the back propagation algorithm. Results are presented for diagnostic faults in an ammonia-water packed distillation column. First, a 6-4-6 network architecture (six input nodes corresponding to the state variables and six output nodes corresponding to the six malfunctions) was chosen based on the minimum root-mean-square-error and mean absolute percentage error;and a maximum value of the Pearson correlation coefficient (C-P). The values of the learning rate, momentum and the gain terms were taken as 0.8, 0.8 and 1.0, respectively. The detection of the designated faults by the network was good. Relative importance of the various input variables on the output variables was calculated based on the partitioning of connection weights which showed that bottoms temperature, overhead composition and overhead temperature are not much affected by the disturbances in feed rate, feed composition and vapor rate in the given range. This resulted in a simplified 3-4-6 net architecture with similar capabilities as the 6-4-6 net thereby reducing the number of computations. (C) 2003 Elsevier B.V. All rights reserved.
The effects of composition and intercritical heat treatment parameters on tensile strength and percentage elongation of Si-Mn TRIP steels were modeled, using a neural network with a feed forward topology and a back pr...
详细信息
The effects of composition and intercritical heat treatment parameters on tensile strength and percentage elongation of Si-Mn TRIP steels were modeled, using a neural network with a feed forward topology and a back propagation algorithm. It was found that a committee of nets models the experimental data more accurately than a single model. The trained network was then applied to a low-carbon low-silicon steel in order to estimate the appropriate heat treatment process conditions. To explain variations in the mechanical properties, the material was subjected to a typical two-stages intercritical annealing and bainitic holding treatment. According to the results of model, tempering of material for a shorter time results in higher tensile strength and percentage elongation values. This behavior was later confirmed by microstructural studies and was attributed to both higher austenite volume fraction and higher martensite content in the samples tempered for a shorter bainitic holding. (C) 2004 Published by Elsevier B.V.
An integrated Artificial Neural Network (ANN) approach to Short-Term Load Forecasting (STLF) is proposed in this paper. Four modules consisting of the Basic ANN, Peak and Valley ANN, Averager and Forecaster and Adapti...
详细信息
An integrated Artificial Neural Network (ANN) approach to Short-Term Load Forecasting (STLF) is proposed in this paper. Four modules consisting of the Basic ANN, Peak and Valley ANN, Averager and Forecaster and Adaptive Combiner form the integrated method for load forecasting. The Basic ANN uses the historical data of load and temperature to predict the next 24 h load, while the Peak and Valley ANN uses the past peak and valley data of load and temperatures, respectively. The Averager captures the average variation of the load from the previous load behaviour, while the adaptive combiner uses the weighted combination of outputs from the Basic ANN and the Forecaster, to forecast the final load. The regression based and time series methods are conceptually incorporated into the ANN to obtain an integrated load forecasting approach. (C) 2004 Elsevier B.V. All rights reserved.
Now a days, image recognition systems have several applications in enormous fields. The use of recognition systems (based on Artificial Neural Networks) as means of predicting medical diagnosis and recommending succes...
详细信息
ISBN:
(纸本)0769521088
Now a days, image recognition systems have several applications in enormous fields. The use of recognition systems (based on Artificial Neural Networks) as means of predicting medical diagnosis and recommending successful treatments has been a highly active research field in past five years. The purpose of this paper is to construct and train an artificial neural network to serve as a knowledge base that can accurately detect Pulmonary Tuberculosis. The first necessary step is to preprocess the different patients MMR's, which consisted of lesions of Tuberculosis and extract the features. Then the extracted features are converted into usable format (gray scale values) and given to neural net for training. It is based on back propagation algorithm.. Then the knowledge base is used to detect a sample collected from a new patient is given as target to recognize it.
Most of different methods and models to predict sea water level require comprehensive exogenous inputs and involve some analysis along with certain assumptions. This paper describes the development of an Artificial Ne...
详细信息
This paper describes a new method to identify the type of fabric weave by using a neural network classifier. The characteristic parameters of the input layer, derived from fabric image, are composed of the Markov rand...
详细信息
This paper describes a new method to identify the type of fabric weave by using a neural network classifier. The characteristic parameters of the input layer, derived from fabric image, are composed of the Markov random field character, the difference between the maximum and the minimum of gray level projections in weft and warp directions, the area ratio of the brightness region to the total area in image, the weft and the warp yarn count. The experimental results show that the neural network classifier can effectively classify fabric weave with 98.33% of accuracy, which is helpful in the recognition of fabric weave parameters.
An automated AT89C55WD microcontroller based surface roughness and distance measurement system is developed. An IR diode and detector are used for the measurement. In this paper, the experimental arrangement for the m...
详细信息
An automated AT89C55WD microcontroller based surface roughness and distance measurement system is developed. An IR diode and detector are used for the measurement. In this paper, the experimental arrangement for the measurement of distance and classification of surface roughness is explained. An artificial neural network (ANN), trained with a back propagation algorithm is employed for the data processing to classify the roughness of the surface and for the distance measurement. A three layer ANN is used to learn the type of surface and corresponding output using a back propagation algorithm. After training, the neural network is used to classify the roughness of the surface and to measure distance.
This paper develops a color image vehicular detection (CIVD) system in which background differencing technique is employed to detect whether a vehicle passes through the detecting points equally spaced out on a pseudo...
详细信息
This paper develops a color image vehicular detection (CIVD) system in which background differencing technique is employed to detect whether a vehicle passes through the detecting points equally spaced out on a pseudo line detector. Two methods (interval search and regression) are tried to determine the optimal crisp threshold values to cope with various lighting conditions. To compare the detection performance with and without incorporating a fuzzy neural network (FNN), a three-layer FNNCIVD system is further designed with trapezoidal membership function and network parameters trained by back propagation algorithm. Under different environments (freeway and urban street) with various lighting conditions (daytime and nighttime), it is found that the detection success rates for interval-search CIVD and regression CIVD are about the same. However, both perform worse than the FNNCIVD system in which about 90% success rates are reported with seven detection points. Compared with the interval-search CIVD system, the FNNCIVD system can increase the success rates at a range of 14% to 22% on the freeway mainline and 18% to 26% on the urban street. It is also found that daytime detection performance is slightly better than nighttime detection. Possible reasons for missed detection and false detection are discussed.
暂无评论