The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
To deal with the problems which exit in ADR (adverse drug reaction) signal detection and automatic warning field in China, the methods which used in ADR signal detection have been analyzed firstly in this paper. In this condition, the algorithm model of ADR signal detection and realization process have been studied by computer based on BCPNN method and SRS database. And then the realization algorithm...
Quality evaluation and classification is very important for crop market price determination. A lot of methods have been applied in the field of quality classification including principal component analysis (PCA) and artificial neural network (ANN) etc. The use of ANN has been shown to be a cost-effective technique. But their training is featured with some drawbacks such as small sample effect, black...
This paper discusses the convergence of completely connected recursive neural networks. That is, we analyze the affection of output of the neural networks under which the error donpsilat accumulate, brought by disturbances of inputs and weights. The corresponding conditions are obtained, which are also the conditions of convergence of the neural networks.
This paper presents a new predistortion scheme based on cascade-correlation (CasCor) neural network for high-power amplifier (HPA) with memory in an orthogonal frequency division multiplexing(OFDM) system. An efficient algorithm to update the neural network weights parameters is derived.Simulation results show that the proposed neural network predistorter not only suppresses out-of-band spectrum emission...
The main objective of this work is to develop a new method for the blind separation of the noised image. A nonlinear neural network and independent component analysis (ICA) algorithm are combined. The neural network filter is used to remove the noise and ICA algorithm is used for the blind separation of the mixed image. But the effect of pre-filter is different from the post-filter. By comparing the...
The singular value eigenvectors of the different kinds of the mineral oil stylebooks are obtained by parameterizing the three-dimensional fluorescence spectroscopy. They are complicated and not easy to be recognized by the simple formula. The multiwavelet neural network is introduced to realize the identification of the different kinds of the mineral oil. It was layered. It had the feature of the...
Neural network techniques have been widely applied to areas of such as data mining, information integration and grid computing. This paper proposes a new learning algorithm based on trust region optimization theory. In the paper, the Dogleg-algorithm to obtain the valid trust region steps is presented, and a self-adjustable method with variable coefficients is given to resolve the problem of oscillatory...
By the analyze of chaos for runoff series, combing the reconstruction phase space theory and BP neural network to develop the BP neural network model based reconstruction phase space, and forecast the runoff series mensal in Xiaoqing river hydrological station of Jinan, the result shows that the model has a very good forecast accuracy and value.
The local linear wavelet neural network is an improvement of wavelet network and commonly used learning algorithm is gradient descent method. In this paper, we attempt to predict sunspots using a local linear wavelet neural network and weight perturbation technique like simulated annealing. The simulation results show the effectiveness of the proposed method.
Parameters and individual ability in Dichotomously Scored of Response Theory model are estimated with Back-Propagation Neural Network. The dimension of Scoring matrixes X is descended by using scoring rate or passing rate or coefficient of correlation or guess rate when estimating those item parameters. The method is simulated in computer, and the results show that the item parameters estimation is...
This paper expounds the principle of BP neural network with applications to image compression and the neural network models. Then an image compressing algorithm based on improved BP network is developed. The blocks of original image are classified into three classes: background blocks, object blocks and edge blocks, considering the features of intensity change and visual discrimination. The BP algorithm...
Electronic tongue is a device which is used to classify different taste by multi-sensor. In this work, we had measured the production of chemical composition of five different mineral water by four kinds of selected ion array (sensitive to H+, Na+, Ca2+ and K+, respectively). Principal component analysis, a kind of multivariate data analysis was used to educing of total number of the sensors in the...
In this paper, we propose a novel artificial neural network (ANN) demodulator to demodulate FSK signal. It has some important features compared with conventional method. Firstly, the anti-interference ability of ANN demodulator is better than that in traditional way. In traditional receiver, there must be a band-pass filter (BPF) to filter out-of-band noise; however, in ANN demodulator, the signal...
With the advent of digital technology, digital image has gradually taken the place of the original analog photograph, and the forgery of digital image has become increasingly easy and indiscoverable. To implement image splicing blind detection, this paper proposes a new splicing detection model. Image splicing detection can be treated as a two-class pattern recognition problem, which builds the model...
Process burning coal powder in boiler is in very complex suspension combustion and very unstable, monitoring system image includes a lot of noise signals from difference hands. In this chapter, compose of furnace flame monitoring system is introduced, the new method called as fast median filter for elimination image noise and special boundary detection algorithm based on neural network technique for...
Chaos and artificial neural networks have been providing a new rout for investigating the complicated nonlinear time series. As the traditional neural networks are easy to get slow convergence and produce large redundancy which might consequently bring low efficiency, the biased wavelet neural networks is employed to build chaotic time series forecasting. Efforts are also made to assess the forecast...
In this paper, a practical digital watermarking system is introduced. The embedding and extraction algorithms of the binary digital watermarking and their realization are described. Some experiment results against attack are given. At last, a practical digital watermarking system based on the spatial domain transform and BP neural network is designed and realized with MATLAB.
This paper is to effectively solve the problem that the objects of traditional plant identification were too broad and the classification features of it were usually not synthetic and the recognition rate was always slightly low. This study gives one recognition approach, in which the shape features and the texture features of the leaves of broad-leaved trees combine, composing a synthetic feature...
In equipment monitoring and fault diagnosis, correctly evaluating and selecting the signal features contribute greatly to the effectiveness and accuracy of recognition result. Because it is difficult to create a criterion to evaluate the feature of measured signals in condition of small samples when we use traditional statistic pattern recognition theory to do this, this paper put forward a method...
It is very important to study non-uniformity correction algorithm in infrared focal plane array (IRFPA). In order to improve the convergence speed and non-stability in traditional neural network non-uniformity correction algorithm, a new scene-based non-uniformity correction algorithm for IRFPA was designed in this paper. The algorithm firstly arrange a pixelpsilas gray value and its around eight...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.