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.
Construction of robust and accurate deep neural networks (DNNs) is a computationally demanding and time-consuming process. Such networks also end up being memory intensive. Today, there is ever-increasing need to provide proactive and personalized support for users of smart devices. We could provide better personalization if we have the ability to update/train the DNN on edge devices. Also, by moving...
Carefully injected noise can speed the convergence and accuracy of video classification with recurrent backpropagation (RBP). This noise-boost uses the recent results that backpropagation is a special case of the generalized expectation maximization (EM) algorithm and that careful noise injection can always speed the average convergence of the EM algorithm to a local maximum of the log-likelihood...
In the proposed work, we presented an Artificial Neural Network approach to predict the stock market indices. We outlined the design of the Neural Network model with its salient features and customizable parameters. A number of the activation functions are implemented along with the options for the cross validation sets. We finally test our algorithm on the Nifty stock index dataset where we predict...
Fast and efficient implementation of elementary functions such as sin(), cos(), and log() are of ample importance in a large class of applications. The state of the art methods for function evaluation involves either expensive calculations such as multiplications, large number of iterations, or large Lookup-Tables (LUTs). Higher number of iterations leads to higher latency whereas large LUTs contribute...
Deep Convolutional Neural Networks (CNN) enforce supervised information only at the output layer, and hidden layers are trained by back propagating the prediction error from the output layer without explicit supervision. We propose a supervised feature learning approach, Label Consistent Neural Network, which enforces direct supervision in late hidden layers in a novel way. We associate each neuron...
Restricted Boltzmann Machine (RBM) is a generative stochastic energy-based model of artificial neural network for unsupervised learning. Recently, RBM is well known to be a pre-training method of Deep Learning. In addition to visible and hidden neurons, the structure of RBM has a number of parameters such as the weights between neurons and the coefficients for them. Therefore, we may meet some difficulties...
A learning process is easily trapped into a local minimum when training multi-layer feed-forward neural networks. An algorithm called Wrong Output Modification (WOM) was proposed to help a learning process escape from local minima, but WOM still cannot totally solve the local minimum problem. Moreover, there is no performance analysis to show that the learning has a higher probability of converging...
Delay learning in SpikeProp is useful because it eliminates the need of redundant synaptic connections in a Spiking Neural Network (SNN). The delay learning enhancement to SpikeProp, however, also inherits the complications present in basic SpikeProp with weight update that obstruct the learning process. To tackle these issues, we perform delay convergence analysis to investigate the conditions required...
Enterprise in financial trouble is a comprehensive event and the enterprise financial situation can be reflected through the liquidity ratio, earnings per share and net assets per share and cash content per share. Artificial neural network method is used to establish the financial early warning model to find the potential financial crisis at an early age. The experiment results show that BP neural...
Recently, the application of complex-valued neural networks (CVNNs) for real-valued classification has attracted more and more attention. To overcome the limitations of the existing CVNNs, this study extends the real-valued group method of data handling (RGMDH) type neural network to complex domain, and constructs complex-valued GMDH-type neural network (CGMDH). First, it proposes the complex least...
In this paper we compare the performance of back propagation and resilient propagation algorithms in training neural networks for spam classification. Back propagation algorithm is known to have issues such as slow convergence, and stagnation of neural network weights around local optima. Researchers have proposed resilient propagation as an alternative. Resilient propagation and back propagation...
In this paper, we discussed respective superiority what back propagation neural network based on fractional differential and integer-order differential have, from two aspectsconvergent speed and error. Then in order to get better convergent effect, the paper proposes the neural network of adaptive order. The detailed progress what is verified by MATLAB is illustrated in figures as follows.
In human brain the neurons are excited in a dynamic way. The response of different neurons varies widely because of the variation of electrical signal in every neuron. Backpropagation(BP) is a training algorithm where the learning of the Neural Network (NN) is done by a constant learning rate (LR). But to replicate the human brain function, the learning rate should be changed as the excitation of...
Hydraulic bending roller is a most basic and important method for shape control of strip. The rolled shape quality is decided by the setting value of bending farce in great part. This paper chooses five-stand hot tandem rolling mill in 1810 product line of Tangshan Iron and Steel Company as background, and deals primarily with the study of the bending force prediction model of the rolling unit. To...
SpikeProp is a supervised learning algorithm for spiking neural networks analogous to backpropagation. Like backpropagation, it may fail to converge for particular networks, parameters and datasets. However there are several behaviours and additional failure modes unique to SpikeProp which have not been explicitly outlined in the literature. These factors hinder the adoption of SpikeProp for general...
In the present work, a change detection technique in remotely sensed images (under the scarcity of labeled patterns) is proposed where an ensemble of semi-supervised classifiers is used, instead of using a single (weak) classifier. Iterative learning of multiple classifier system is carried out using the selected unlabeled patterns along with a few labeled patterns. Selection of unlabeled patterns...
Increased demands for higher storage capacity solution have driven the Hard Disk Drive (HDD) technological boundaries. As the Perpendicular Magnetic Recording (PMR) head shows promising increase in Areal Density away from the limit of Longitudinal Magnetic Recording, HDD companies have switch to 100% PMR drives. PMR heads requires tight physical specifications fabricating its Writer Element in order...
The conventional algorithm of the BP neural network has some disadvantages such as in the vicinity of the target, if the learning factor is too small, the convergence may be too slow, and if the learning factor is too large, the convergence may be amended too much, leading to oscillations and even dispersing phenomenon. At the same time, the very slow speed of convergence and the main procedure is...
In this paper, we propose a method dealing with the problem of image restoration based on the conception of Back propagation Neural Network Algorithm. Conventional Back Propagation Algorithm has its inherited drawbacks, i.e. slow convergence rate, long training time, hard to achieve global minima etc. Recently, several methods introduced the dynamic learning rate and the dynamic momentum coefficient...
We present a generative model for a three-layer hierarchy consisting of a retinal layer, a simple-cell layer, and a complex-cell layer. The weights in the model are trained using supervised learning on the retinal layer and complex cells. Once the weights are learned, the model is able to perform bottom-up classification of images, as well as top-down reconstruction from a specified category. The...
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.