Search

Hand gestures recognition goals can only be fulfilled when gesture isolation is coupled with an effective feature extraction followed by highly efficient classification. In the context of machine vision, feature extraction and classification can be jointly called pattern recognition in which, previous known patterns are matched with a query gesture. [ABSTRACT FROM AUTHOR]

Insulated gate bipolar transistors (IGBT) are used in critical application areas such as inverters in hybrid cars, motion control systems for variable speed motor drives and high power switch mode power supplies. In these critical applications, diagnostic approaches are required to monitor the health and predict the reliability of these devices to prevent system downtime and costly failures. In this study, Mahalanobis distance (MD) was used for anomaly detection for non-punch through (NPT) IGBTs. The IGBTs were aged by electrical-thermal stress under a resistive load until their failure. Monitored collector-emitter voltage and collector-emitter currents were used as input parameters to calculate the MD. The MD values obtained from the healthy data were transformed using a Box–Cox transform and three sigma limits were obtained from the transformed data. The upper three sigma limit of the transformed MD healthy data was used as a threshold to detect degradation in the IGBTs. The MD anomaly detection approach allowed for detection time of anomalies in NPT IGBTs up to 77% before failure. [ABSTRACT FROM AUTHOR]

In this paper, we propose an iterative self-organizing map approach for spatial outlier detection (IterativeSOMSO). IterativeSOMSO method can address high dimensional problems for spatial attributes and accurately detect spatial outliers with irregular features. Detection of spatial outliers facilitates further discovery of spatial distribution and attribute information for data mining problems. The experimental results indicate our proposed approach can be effectively implemented for the large spatial dataset based on U.S. Census Bureau with approving performance. [ABSTRACT FROM AUTHOR]

k-nearest neighbor method (kNN) is a very useful and easy-implementing method for real applications. The query point is estimated by its k nearest neighbors. However, this kind of prediction simply uses the label information of its neighbors without considering their space distributions. This paper proposes a novel kNN method in which the centroids instead of the neighbors themselves are employed. The centroids can reflect not only the label information but also the distribution information of its neighbors. In order to evaluate the proposed method, Euclidean distance and Mahalanobis distance is used in our experiments. Moreover, traditional kNN is also implemented to provide a comparison with the proposed method. The empirical results suggest that the propose method is more robust and effective. [ABSTRACT FROM AUTHOR]

Many learning algorithms rely on distance metrics to receive their input data. Research has shown that these metrics can improve the performance of these algorithms. Over the years an often popular function is the Euclidean function. In this paper, we investigate a number of different metrics proposed by different communities, including Mahalanobis, Euclidean, Kullback-Leibler and Hamming distance. Overall, the best-performing method is the Mahalanobis distance metric. [ABSTRACT FROM AUTHOR]

Kernel methods for data analysis are frequently considered to be restricted to positive definite kernels. In practice, however, indefinite kernels arise e.g. from problem-specific kernel construction or optimized similarity measures.We, therefore, present formal extensions of some kernel discriminant analysis methods which can be used with indefinite kernels. In particular these are the multi-class kernel Fisher discriminant and the kernel Mahalanobis distance. The approaches are empirically evaluated in classification scenarios on indefinite multi-class datasets. [ABSTRACT FROM AUTHOR]

Neural Networks (NNs) have been widely used in many industrial processes for prediction and optimization and they have been proven to be useful tools for explaining complex processes. The main objective of this work consists of improving the accuracy of a Radial Basis Function Neural Network Redesigned by Genetic Algorithm and Mahalanobis distance for predicting a welding process. The evaluation function in this approach considers the use of the Coefficient of Determination R2. The results indicated that the statistical method R2 is a good alternative to validate the efficiency of the Neural Network model. The principal conclusion in this work is that the Radial Basis Function Redesigned by Genetic Algorithm and Mahalanobis distance had a very good performance in a real case, considering the prediction of specific responses in a welding process. [ABSTRACT FROM AUTHOR]

For its simplicity and applicability, fuzzy c-means clustering algorithm is widely used in image segmentation. However, fuzzy c-means clustering algorithm has some problems in image segmentation, such as sensitivity to noise, local convergence, etc. In order to overcome the fuzzy c-means clustering shortcomings, this paper replaces Euclidean distance with Mahalanobis distance in the fuzzy c-means clustering algorithm. Experimental results show that the proposed algorithm has a significant improvement on the effect and efficiency of segmentation comparing with the standard FCM clustering algorithm. [ABSTRACT FROM AUTHOR]

Types of files (text, executables, Jpeg images, etc.) can be identified through file extension, magic number, or other header information in the file. However, they are easy to be tampered or corrupted so cannot be trusted as secure ways to identify file types.In the presence of adversaries, analyzing the file content may be a more reliable way to identify file types, but existing approaches of file type analysis still need to be improved in terms of accuracy and speed. Most of them use byte-frequency distribution as a feature in building a representative model of a file type, and apply a distance metric to compare the model with byte-frequency distribution of the file in question. Mahalanobis distance is the most popular distance metric. In this paper, we propose 1) the cosine similarity as a better metric than Mahalanobis distance in terms of classification accuracy, smaller model size, and faster detection rate, and 2) a new type-identification scheme that applies recursive steps to identify types of files. We compare the cosine similarity to Mahalanobis distance using Wei-Hen Li et al.΄s single and multi-centroid modeling techniques, which showed 4.8% and 13.10% improvement in classification accuracy (single and multi-centroid respectively). The cosine similarity showed reduction of the model size by about 90% and improvement in the detection speed by 11%. Our proposed type identification scheme showed 37.78% and 31.47% improvement over Wei-Hen Li΄s single and multi-centroid modeling techniques respectively. [ABSTRACT FROM AUTHOR]

An intelligent frog call identification agent is developed in this work to provide the public to easily consult online. The raw frog call samples are first filtered by noise removal, high frequency compensation and discrete wavelet transform techniques in order. An adaptive end-point detection segmentation algorithm is proposed to effectively separate the individual syllables from the noise. Four features are extracted and serve as the input parameters of the classifier. Three well-known classifiers, the k-th nearest neighboring, Support Vector Machines and Gaussian Mixture Model, are employed in this work for comparison. A series of experiments were conducted to measure the outcome performance of the proposed agent. Experimental results exhibit that the recognition rate for Gaussian Mixture Model algorithm can achieve up to the best performance. The effectiveness of the proposed frog call identification agent is thus verified. [ABSTRACT FROM AUTHOR]

Distance-based record linkage (DBRL) is a common approach to empirically assessing the disclosure risk in SDC-protected microdata. Usually, the Euclidean distance is used. In this paper, we explore the potential advantages of using the Mahalanobis distance for DBRL. We illustrate our point for partially synthetic microdata and show that, in some cases, Mahalanobis DBRL can yield a very high re-identification percentage, far superior to the one offered by other record linkage methods. [ABSTRACT FROM AUTHOR]