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.
Deep multi-layer neural networks are generally trained using variants of the gradient descent based algorithm. However, this kind of algorithms usually encounter a series of shortcomings, such as low training efficiency, local minimum, difficult control parameter tuning, and gradient vanishing or exploding. Besides, for a specific application, how to design the structure of the network, that is, how...
The success of deep learning proves that deep models are able to achieve much better performance than shallow models in representation learning. However, deep neural networks with auto-encoder stacked structure suffer from low learning efficiency since common used training algorithms are variations of iterative algorithms based on the time-consuming gradient descent, especially when the network structure...
Deep learning scheme has received significant attention during these years, particularly as a way of building hierarchical representations from unlabeled data for a variety of signal and information processing tasks. However, deep neural networks suffer from slow learning speed since most used training algorithms are based on variations of the gradient descent algorithms which require iterative optimization...
Brill tagging is a classic rule-based algorithm for part-of-speech (POS) tagging that assigns tags, such as nouns, verbs, adjectives, etc., to input tokens. Due to the the intense memory requirements of rule matching, CPU implementations of the Brill tagging algorithm have been found to be slow. We show that Micron's Automata Processor (AP) — a new computing architecture that can perform massively...
Considering stability problems of the current corridor recognition algorithm based on laser scanner, this paper proposes a method for mobile robot fast laser-based corridor recognition. Firstly, Median filter is used to deal with the laser scan to reduce the noise. Then, it employs segmentation method to cluster the data to remove some isolated points. After that, it is feasible to calculate the direction...
Identifying bad objects hidden amidst many good objects is important for public safety and decision-making. These problems are complicated in that the cost of leaving a bad object unidentified may not be specified easily, making it difficult to apply existing cost-sensitive classification that depends on knowing a cost matrix or cost distribution. A compelling case for this "illusive cost"...
In facial expression recognition, high dimensional feature processing is still a hot topic since the solution to this problem can considerably reduce the time consuming operation and computational memory. Many methods have been developed to reduce feature dimension and extract the fundamental information in the feature space by projecting the original data into some lower dimensional space. In this...
The fault identification of power system is of great significance in the event of failure This paper introduce a fault identification method based on multi-wavelet packet and artificial neural network. Firstly, through the simulation of a two-500Kv power source transmission line on PSCAD/EMTDC, the variety of fault signals is generated in different conditions. Then, these fault signals are decomposed...
Two-dimensional Principal component analysis (2DPCA) is widely used in face feature extraction and recognition as its lower-computational complexity comparing with principal component analysis (PCA). In this paper, we propose a feature extraction algorithm of pulmonary nodules based on 2DPCA with adaptive parameters. The cumulative variance proportion which is the histogram peak value of CT image...
Detection of pulmonary nodules combined of extraction by multi-directions PCA and identification by 3D (three dimension) BP neural network is presented in the paper, which is different from most lung CAD algorithms, that it does not require any a priori information by human intervention but solely the information contained by the CT image itself, and it is capable to perform full automation which...
The study concerns with classification of acoustic emission signals in composite laminates using support vector machine (SVM). Wavelet packet analysis is performed initially to extract the features and to reduce the dimensionality of original data features. The SVM classifiers are trained with a subset of the experimental data for known fault conditions and are tested using the remaining set of data...
This paper analyses the characteristics of information flow in military supply chain, the influencing factors and the structure of military supply chain management. Then, the method of information flow management and control in military supply chain is designed, and the architecture of information flow control in military supply chain management information system is proposed.
In this paper, a novel digital watermarking scheme is devised based on improved Back-Propagation neural network (BPN) for color image. The watermark is embedded into the discrete wavelet domain of the original image and extracted by training BPN, which can learn the characteristic of the image. For improving the performance of traditional BPN, we consider the adding of momentum coefficient to reduce...
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.