When using Neural Networks (NN) to handle class imbalance problems, there exists a fact that minority class makes less contribution to the error function than the majority class, so the network learned prefers to reco...
详细信息
ISBN:
(纸本)9783642033476
When using Neural Networks (NN) to handle class imbalance problems, there exists a fact that minority class makes less contribution to the error function than the majority class, so the network learned prefers to recognizing majority class data which we pay less attention to. This paper proposes a novel algorithm WNN (Weighted NN) to solve this problem using a newly defined error function in bp (bp) algorithm. Experimental results executed on 20 UCI datasets show that the approach can effectively enhance the recognition rate of minority class data.
Analysing vibration signal is an effective important method for diesel engine fault diagnosis, and its key techniques are feature extraction and pattern recognition. In this paper, wavelet packet decomposition algorit...
详细信息
ISBN:
(纸本)9781424458479
Analysing vibration signal is an effective important method for diesel engine fault diagnosis, and its key techniques are feature extraction and pattern recognition. In this paper, wavelet packet decomposition algorithm as an effective method for fault feature extraction is used to decompose the vibration signals, and its percentage of energy band wavelet packet and wavelet packet energy spectrum entropy are regarded as diagnostic feature vectors. At the same time, in the process of pattern recognition, a mixed neural network training algorithm GA-bp algorithm was used to recognize the fault pattern in fault diagnosis of valve gap abnormal fault. This method can effectively and reliably be used in the fault diagnosis of valve gap abnormal fault by comparing the two algorithms and analyzing the results of real examples. This method can also effectively be used in other fields.
This The pellet sintering process is a physical chemistry process, which is pure time-delay and non-linear with more variables and discrete parameters. In order to solve this problem, a new bp algorithm of NN for the ...
详细信息
ISBN:
(纸本)9781424458479
This The pellet sintering process is a physical chemistry process, which is pure time-delay and non-linear with more variables and discrete parameters. In order to solve this problem, a new bp algorithm of NN for the temperature identification of sintering shaft furnace is introduced in this paper. bp network is combined with PID control and simulation effects are given with MATLAB. The experiment results show this method can make the temperature of firebox conform the requirement of the control standard and can also improve the output and quality of the products.
bp network is the most widely used of the neural net work model, but there are many problems of slow convergence speed and easily getting into the local minimum in the conventional bp algorithm. For this, an improved ...
详细信息
ISBN:
(纸本)9783037852828
bp network is the most widely used of the neural net work model, but there are many problems of slow convergence speed and easily getting into the local minimum in the conventional bp algorithm. For this, an improved algorithm is proposed. Momentum term is added, steepness factors are introduced and adaptive learning rate adjustment factor is added. In the Matlab platform simulations are carried out by each improvement methods on the same bp neural network. The results show that: Convergence of improved bp network is decreased from 1000 to 49 and the error is decreased from 10(-2) to 10(-6). The convergence speed has been significantly improved and the error has been decreased. Using the synthesis improvement method effect is obvious and it provides a good theoretical basis for the practical application.
This paper proposed a new method combining infrared and visible image, which is devoted to monitor the electrical equipment. The key technologies of infrared and visible image processing include two aspects. On the on...
详细信息
ISBN:
(纸本)9781509012107
This paper proposed a new method combining infrared and visible image, which is devoted to monitor the electrical equipment. The key technologies of infrared and visible image processing include two aspects. On the one hand, it researches on targeting localization based on visible image. We can get the horizontal and vertical adjustments of the refitted cradle head camera through the coordinates of the electrical equipment in the image by means of the strong ability for bp algorithm fitting nonlinear function;on the other hand, the infrared image is processed by qualitative method based on the temperature difference The improved MLP network eliminated the harmful effect of the size of the partial derivative on the weight step. The network received some features as the inputs and classified the conditions of equipment into two categories which are defective and normal. Experimental results show that the positioning error is less than 5 pixels and the classification accuracy rate reaches 84.21%, which means that the method not only improves the measurement efficiency of the infrared detection technology, which is not affected by the limitation of measurement environment.
The reinforcement rib design is one of the key parts in entire bottle design. This paper presents the rib performance prediction system based on the bp algorithm and the finite element analysis, which adopts the finit...
详细信息
ISBN:
(纸本)9783037852590
The reinforcement rib design is one of the key parts in entire bottle design. This paper presents the rib performance prediction system based on the bp algorithm and the finite element analysis, which adopts the finite element analysis results as its learning samples, sets up the rib performance prediction system with bp artificial neural network. The results show that the artificial neural network plays an important role in rib performance prediction;meanwhile it can guide the bottle design in practical terms.
Music signal spectrum visualization is of great significance in music processing and music information retrieval. In this study, the application effect of the improved RPROP method and bp algorithm in music signal spe...
详细信息
First, the background, significance and general implementation of vehicle license plate recognition (VLPR) are introduced. Based on analyzing the theory of neural network pattern recognition system, toward the limitat...
详细信息
ISBN:
(纸本)9780769536156
First, the background, significance and general implementation of vehicle license plate recognition (VLPR) are introduced. Based on analyzing the theory of neural network pattern recognition system, toward the limitation of standard bp, this thesis offers its improved method and a real recognition example. Finally, conclusion on its characteristic is given. The improved bp algorithm had some advantages in reaching high error precision, fast convergence speed, short recognition time and high recognition rate.
In the rapid development of neural network prediction today, the use of big data algorithm means to improve the CFG pile composite foundation bearing capacity has become a trend, in order to be more adaptive to today&...
详细信息
Ux interaction design, which originated at the intersection of computer science and cognitive psychology, aims to optimize the interaction between humans and machines to provide a more intuitive and satisfying experie...
详细信息
暂无评论