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.
The paper presents a Time Of Arrival (TOA) location algorithm for Ultra-Wideband (UWB) wireless communication based on Back Propagation (BP) neural network (NN). The following six modules are employed to simulate the locating under Additive White Gaussian Noise (AWGN) channel model: generating samples for training with PPM-TH-UWB signal, training the BP neural network, generating another samples for...
In this paper, a novel supervised architecture for binary classification based on local modelling and information theory is described. The architecture is composed of two steps: in the first one, a separating borderline between the two classes is piecewise constructed by a set of centroids calculated by a modified clustering algorithm, based on information theory; each of these centroids define a...
Learning Vector Quantization (LVQ) is a popular class of nearest prototype classifiers for multiclass classification. Learning algorithms from this family are widely used because of their intuitively clear learning process and ease of implementation. They run efficiently and in many cases provide state of the art performance. In this paper we propose a modification of the LVQ algorithm that addresses...
Fingerprint-Spectrum (FS) -Similitude (FSS) -Estimation (FSSA) is a key stage of medicine services and manages, which decides the truth, reliability, accuracy, safety, and availability of a medicine product. During developing this method people have meat problems such as too large data-volume, 2-dimensional (x- and y-axis) -float, complicated disturbance, and dependence on human-intervening, facing...
How to choose a proper number of the neighbors is an important issue of the locally linear embedding algorithm. To investigate this issue, we propose an optimized locally linear embedding algorithm with adaptive neighbors (ANLLE). The ANLLE selects the neighbors with a locally adaptive criterion. In addition, a new data point mapping method that computes the low-dimensional description of the correspondents...
A new algorithm, Laplacian minmax discriminant projection (LMMDP), is proposed in this paper for supervised dimensionality reduction. LMMDP aims at learning a linear transformation which is an extension of linear discriminant analysis (LDA). Specifically, we define the within-class scatter and the between-class scatter using similarities which are based on pairwise distances in sample space. After...
In this paper we develop a voice activity detection algorithm based on spectrum estimation of speech and non-speech segments using vector quantization method. In this method, we try to classify entry speech signal to speech and non-speech classes. Commonly, the performance of the voice activity detection (VAD) algorithms in non-stationary background noise is not so satisfying under low SNR, so we...
By analyzing and improving the artificial immune algorithm based on information entropy and Euclidean distance, a novel artificial immune algorithm used in intrusion detection (AIAID) is proposed. The core of the algorithm lies on improving the calculation about similarity and expected breed rate by the principle of Mahalanobis distance, namely introducing the importance and value range of antibody...
A novel touching cells splitting algorithm by using concave points and ellipse fitting is proposed to split circle-like or ellipse-like touching cells in this paper. The algorithm is divided into two parts. The first part is contour pre-processing, whose purpose is to find the concave points of the contour and separate the contour into different segments using the concave points. The second part is...
This article presents a method for the calculating similarity of two trajectories. The method is especially designed for a situation where the points of the trajectories are distributed sparsely and at non-equidistant intervals. The proposed method is based on giving different weights to different points: points that are close to each other get smaller weights than the points that do not have neighbors...
Diverse pose estimation of three-dimensional (3D) object in the whole view-space remains a challenge in the field of pattern recognition. In this paper, a pose estimation algorithm of 3D object named isomap-eigenanalysis-regression (Isomap-E-R), which estimates arbitrary pose of 3D object in the whole view space, is proposed. For the training set, the low-dimensional embedding of input pattern set...
This paper presents a new algorithm named kernel bisecting k-means and sample removal (KBK-SR) as a sampling preprocessing for SVM training to improve the scalability. The novel clustering approach kernel bisecting k-means in the KBK-SR tends to fast produce balanced clusters of similar sizes in the kernel feature space, which makes KBK-SR efficient and effective for reducing training samples for...
In this paper we evaluate k-nearest neighbor (KNN), linear and quadratic discriminant analysis (LDA and QDA, respectively) for embedded, online feature fusion which poses strong limitations on computing resources and timing. These algorithms are implemented on our multisensor data fusion (MSDF) architecture and are applied to traffic monitoring, i.e., classifying vehicles using distributed image,...
Nearest neighbor is the basic method in instance-based learning, which is used to approach the real and discrete objective function. In order to enhance the learning speed in nearest neighbor, the optimization of KD tree algorithm was applied in the nearest neighbor method by building the index of the training set. Proper adjustments of the inserting order of the training set can bring the tree more...
Self-organizing map (SOM) has been recognized as a powerful tool in cluster analysis. This paper presents a fuzzy SOM algorithm for mixed numeric and categorical data which integrates fuzzy set theory in model exploration through a fuzzy projection instead of crisp projection. In addition, a hybrid clustering approach is proposed combining SOMs with partitive clustering algorithms for the sake of...
The goal of our research is to develop a grammar induction system that can assign descriptive sentences to ontology models represented by an extended conceptual graph (ECG) which is a conceptual modeling language for describing the semantics of an agentpsilas internal knowledge model. In the proposed system, the ECG model of the agent is converted into a symbolic language sentence. For this, the system...
In this paper I present an environment and algorithm for lazy (incremental) construction of multigram profile models as part of IR (information retrieval) training and exploitation processes. N-grams are traditionally used for natural language text models, but they can be also successfully used for domain independent document classification. I am demonstrating results of a prototype utility which...
This paper presents the logical structure and physical design of a mixed software-hardware system developed for fast codebook generation using the well known LBG algorithm. The system uses a neurocoprocessor based on a Self Organization Feature Map (SOM) in order to make a HW-SW partition of algorithm tasks. The clustering task, the most time demanding, is carried out by hardware, exploiting the intrinsic...
Money laundering (ML) is a serious crime which makes it necessary to develop detection methods in transactions. Some researches have been carried on, but the problem is not thoroughly solved. Aiming at the low detection rate of suspicious transaction at home and abroad in financial field, and with an analysis of radial basis function (RBF) neural network, we propose a radial basis function neural...
A new algorithm of Web text clustering mining is presented, which is based on the Discovery Feature Sub-space Model (DFSSM). This algorithm includes the training stage of SOM and the clustering stage, which characterizes self-stability and powerful antinoise ability. It can distinguishes the most meaningful features from the Concept Space without the evaluation function. we have applied the algorithm...
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.