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.
In the context of a syntactic approach to pattern recognition, there have been several studies in the last few decades on theoretical models for generating or recognizing two-dimensional objects, pictures and picture languages. Motivated by these studies we introduce a new notion of recognizability for a class of picture languages called iso picture languages. We first introduce a notion of local...
As shown in the bibliography, training an ensemble of networks is an interesting way to improve the performance with respect to a single network. However there are several methods to construct the ensemble and there are no complete results showing which one could be the most appropriate. In this paper we present a comparison of eleven different methods. We have trained ensembles of 3, 9, 20 and 40...
One problem in the field of machine learning is that the performance on the training and validation sets lack robustness when applied in real-life situations. Recent advances in ensemble methods have demonstrated that robust behavior can be improved by combining a large number of weak classifiers. The key insight of this paper is that the performance enhancement due to combining multiple classifiers...
This paper shows, through experimental results, that artificial neural networks are good classifiers for the text categorization task. The paper compares the results of experiments on text categorization using Multilayer Perceptron, Self-organizing Maps, C4.5 decision tree and PART decision rules. The experiments were carried out with K1 collection of web documents.
Most computational models of gender classification use global information (the full face image) giving equal weight to the whole face area irrespective of the importance of the internal features. Here we use a two-way representation of face images that includes both global and featural information. We use dimensionality reduction techniques and a support vector machine classifier and show that this...
This paper presents a new symbolic classifier based on a region oriented approach. Concerning the learning step, each class is described by a region (or a set of regions) in Rp defined by the convex hull of the objects belonging to this class. In the allocation step, the assignment of a new object to a class is based on a dissimilarity matching function which compares...
This paper presents a clustering method for interval-valued data using a dynamic cluster algorithm with adaptive squared Euclidean distances. This method furnishes a partition and a prototype to each cluster by optimizing an adequacy criterion that measures the fitting between the clusters and their representatives. To compare a class with its representative, the method uses an adaptive version of...
A two-pass approach to pattern recognition has been described here. In this approach, an input pattern is classified by refining possible classification decisions obtained through coarse classification of the same. Coarse classification here is performed to produce a group of possible candidate classes by considering the entire input pattern, whereas the finer classification is performed to select...
We address the problem of recognition and retrieval of relatively weak industrial signal such as Partial Discharges (PD) buried in excessive noise. The major bottleneck being the recognition and suppression of stochastic pulsive interference (PI) which has similar time-frequency characteristics as PD pulse. Therefore conventional frequency based DSP techniques are not useful in retrieving PD pulses...
This paper presents a novel approach to combine a neural network based auto-associator and a classifier for the classification of microcalcification patterns. It explores the auto-associative and classification abilities of a neural network approach for the classification of microcalcification patterns into benign and malignant using 14 image structure features. The proposed technique used combination...
In this paper, we consider the problem of time series classification. Using piecewise linear interpolation various novel kernels are obtained which can be used with Support vector machines for designing classifiers capable of deciding the class of a given time series. The approach is general and is applicable in many scenarios. We apply the method to the task of Online Tamil handwritten character...
In this paper, we report the results of recognition of online handwritten Tamil characters. We experimented with two different approaches. One is subspace based method wherein the interactions between the features in the feature spate are assumed to be linear. In the second approach, we investigated an elastic matching technique using dynamic programming principles. We compare the methods to find...
A recognition scheme for handwritten basic Bangla (an Indian script) characters is proposed. No such work has been reported before on a reasonably large representative database. Here a moderately large database of Bangla handwritten character images is used for the recognition purpose. A handwritten character is composed of several strokes whose characteristics depend on the handwriting style. The...
We present a comparison of elastic matching schemes for writer dependent on-line handwriting recognition of isolated Tamil characters. Three different features are considered namely, preprocessed x-y co-ordinates, quantized slope values, and dominant point co-ordinates. Seven schemes based on these three features are compared using an elastic distance measure. The comparison is carried out in terms...
This paper describes development of the automated industry and occupation coding system for the Korean Census records. The purpose of the system is to convert natural language responses on survey questionnaires into corresponding numeric codes according to standard code book from the Census Bureau. We employ kNN(k Nearest Neighbors)-based document classification method and information retrieval techniques...
In this paper, a method for detecting possible presence of abnormality in the endoscopic images of lower esophagus is presented. The pre-processed endoscopic color images are segmented using color segmentation based on 3σ-intervals around mean RGB values. The zero-crossing method of edge detection is applied on the gray scale image corresponding to the segmented image. For the large contours, the...
This paper presents a new fault diagnosis method for industrial images based on a Min-Max Modular (M3) neural network and a Gaussian Zero-Crossing (GZC) function. The most important advantage of the proposed method over existing approaches such as radial-basis function network and support vector machines is that our classifier has locally tuned response characteristics and the misclassification rate...
In this paper, we present the design and application of a pattern classifying machine (PCM) for distributed data mining (DDM) environment. The PCM is based on a special class of sparse network referred to as Cellular Automata (CA). The desired CA are evolved with an efficient formulation of Genetic Algorithm (GA). Extensive experimental results with respect to classification accuracy and memory overhead...
This paper proposes NeurAge, an agent-based multi-classifier system for classification tasks. This system is composed of several neural classifiers (called neural agents) and its main aim is to overcome some drawbacks of multi-classifier systems and, as a consequence, to improve performance of such systems.
MDS algorithms are data analysis techniques that have been successfully applied to generate a visual representation of multivariate object relationships considering only a similarity matrix. However in high dimensional spaces the concept of proximity become meaningless due to the data sparsity and the maps generated by common MDS algorithms fail often to reflect the object proximities. In this...
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.