International Journal of Pattern Recognition and Artificial Intelligence

Feature ranking is widely employed to deal with high dimensionality in text classification. The main advantage of feature ranking methods is their low cost and simple algorithms. However, they suffer from some drawbacks which cause low performance compared to wrapper approach feature selection methods. In this paper, three major drawbacks of feature ranking methods are discussed. First, we show that feature ranking methods are highly problem dependent. For designing an effective feature ranking method and appropriate ranking threshold, we need background knowledge including the data set characteristics as well as the classifier to be used. Second, the feature ranking methods are univariate functions, while the nature of text classification is multivariate. It means that in these methods, correlation between terms is ignored. Finally, they fail in multiple class problems with unbalanced class distribution because they pay more attention to the simpler and larger classes. In this paper, these drawbacks, especially the last two issues, are experimentally investigated using a set of extensive numerical experiments with several data sets and feature scoring measures.

The goal of nonlinear dimensionality reduction is to find the meaningful low dimensional structure of the nonlinear manifold from the high dimensional data. As a classic method of nonlinear dimensional reduction, locally linear embedding (LLE) is more and more attractive to researchers due to its ability to deal with large amounts of high dimensional data and its noniterative way of finding the embeddings. However, several problems in the LLE algorithm still remain open, such as its sensitivity to noise, inevitable ill-conditioned eigenproblems, the inability to deal with the novel data, etc. The existing extensions are comprehensively reviewed and discussed classifying into different categories in this paper. Their strategies, advantages/disadvantages and performances are elaborated. By generalizing different tactics in various extensions related to different stages of LLE and evaluating their performances, several promising directions for future research have been suggested.

Arabic script is naturally cursive and unconstrained and, as a result, an automatic recognition of its handwriting is a challenging problem. The analysis of Arabic script is further complicated in comparison to Latin script due to obligatory dots/stokes that are placed above or below most letters. In this paper, we introduce a new approach that performs online Arabic word recognition on a continuous word-part level, while performing training on the letter level. In addition, we appropriately handle delayed strokes by first detecting them and then integrating them into the word-part body. Our current implementation is based on Hidden Markov Models (HMM) and correctly handles most of the Arabic script recognition difficulties. We have tested our implementation using various dictionaries and multiple writers and have achieved encouraging results for both writer-dependent and writer-independent recognition.

In this paper we propose a thinning methodology applicable to character images. It is novel in terms of its ability to adapt to local character shape while constructing the thinned skeleton. Our method does not produce many of the distortions in the character shapes which normally result from the use of existing thinning algorithms. The proposed thinning methodology is based on the medial axis of the character. The skeleton has a width of one pixel. As a by-product of our thinning approach, the skeleton also gets segmented into strokes in vector form. Hence further stroke segmentation is not required. We have conducted experiments with printed and handwritten characters in several scripts such as English, Bengali, Hindi, Kannada and Tamil. We obtain less spurious branches compared to other thinning methods. Our method does not use any kind of post processing.

In this paper, a facial expression recognition system based on cerebella model articulation controller with a clustering memory (CMAC-CM) is presented. Firstly, the facial expression features were automatically preprocessed and extracted from given still images in the JAFFE database in which the frontal view of faces were contained. Next, a block of lower frequency DCT coefficients was obtained by subtracting a neutral image from a given expression image and rearranged as input vectors to be fed into the CMAC-CM that can rapidly obtain output using nonlinear mapping with a look-up table in training or recognizing phase. Finally, the experimental results have demonstrated recognition rates with various block sizes of coefficients in lower frequency and cluster sizes of weight memory. A mean recognition rate of 92.86% is achieved for the testing images. CMAC-CM takes 0.028 seconds for test image in testing phase.

Distance metric learning is a powerful tool to improve performance in classification, clustering and regression tasks. Many techniques have been proposed for distance metric learning based on convex programming, kernel learning, dimension reduction and large margin. The recently proposed large margin nearest neighbor classification (LMNN) improves the performance of k-nearest neighbors classification (k-nn) by a learned global distance metric. However, it does not consider the locality of data distributions. We demonstrate a novel local distance metric learning method called hierarchical distance metric learning (HDM) which first builds a hierarchical structure by grouping data points according to the overlapping ratios defined by us and then learns distance metrics sequentially. In this paper, we combine HDM with LMNN and further propose a new method named hierarchical distance metric learning for large margin nearest neighbor classification (HLMNN). Experiments are performed on many artificial and real-world data sets. Comparisons with the traditional k-nn and the state-of-the-art LMNN show the effectiveness of the proposed HLMNN.

In this paper, we propose a novel approach to frequency segmentation using wavelet decomposition of pseudo-motion image. In this way, a fixed image is translated such that a sequence of moving images is produced, which are called the 0 pseudo-motion images. In fact, we can consider the translation of a function to be a motion of eyeshot. When a function is translated, its wavelet coefficients will oscillate. From this property, we can detect the special areas of the image. Some experiments were conducted, which are used to find the position of license plates of vehicles in digital images. The experimental results demonstrate the performance of the method.

Co-training is a multiview semi-supervised learning algorithm to learn from both labeled and unlabeled data, which iteratively adopts a classifier trained on one view to teach the other view using some confident predictions given on unlabeled examples. However, as it does not examine the reliability of the labels provided by classifiers on either view, co-training might be problematic. Even very few inaccurately labeled examples can deteriorate the performance of learned classifiers to a large extent. In this paper, a new method named robust co-training is proposed, which integrates canonical correlation analysis (CCA) to inspect the predictions of co-training on those unlabeled training examples. CCA is applied to obtain a low-dimensional and closely correlated representation of the original multiview data. Based on this representation the similarities between an unlabeled example and the original labeled examples are determined. Only those examples whose predicted labels are consistent with the outcome of CCA examination are eligible to augment the original labeled data. The performance of robust co-training is evaluated on several different classification problems where encouraging experimental results are observed.

The laser ranging has extensive detection range and high measurement accuracy and so it is widely applied in navigation systems of the autonomous robot. But different environmental conditions have different influences on the transmission of the laser, producing errors of various degrees to measurement. The proposed range of influence in research on laser ranging LMS200 which is used in robot navigation under dissimilar conditions, by analyzing the primary factor that affects LMS200 ranging in a room and outside, contributing to the experimental contrast, obtaining experimental effects under environmentally influencing factors, discovering reflectivity, incidence, mixed pixel and visibility and so on that are environmental factors truly affecting practical application, builds the foundation for carrying out the ranging application of robot navigation. Mobile robots need enough sensor information of movements for carrying out autonomous navigation. The information establishment environment model in the completely unknown environment, depends upon itself to provide sensors to the robot that carries out autonomous localization and navigation. This paper also systematically introduces the applied research of laser ranging in mobile robot autonomous navigation, carrying out the scanning of the environment through the two-dimensional range finder sensor, proposed in autonomous navigation of map building and localization that uses two-dimensional scanning to obtain three-dimensional data stream, the experiment confirms that this algorithm allows the robot to obtain a very good three-dimensional visual structure drawing.

CAN (Controller Area Network) operation observes event triggering standards, it is widely applied in real-time distributional control system of the automobile electronic control system, satellite control system, medical equipment's electronic control system as well as the textile equipment and so on. As the extension of offline bin-packing problem, periodic task scheduling has many important applications in real-time distributed systems. For time triggered CAN control systems, two strategies to determine the basic cycle (BC) of TTCAN are presented. Next-fit algorithm, next-fit descending algorithm, best-fit algorithm and best-fit descending algorithm are proposed to construct periodic task scheduling tables. Time complexity and worst-case asymptotic performance ratio of these algorithms are analyzed, and the experimental results indicate that NFA algorithm is not worse than NFDA algorithm, BFA is also not worse than BFDA algorithm, four algorithms have the advantage over the typical one-dimensional bin-packing algorithm.