Nearest Mean Classifier Research Articles

Quantum Minimum Distance Classifier

We propose a quantum version of the well known minimum distance classification model called Nearest Mean Classifier (NMC). In this regard, we presented our first results in two previous works. First, a quantum counterpart of the NMC for two-dimensional problems was introduced, named Quantum Nearest Mean Classifier (QNMC), together with a possible generalization to any number of dimensions. Secondly, we studied the n-dimensional problem into detail and we showed a new encoding for arbitrary n-feature vectors into density operators. In the present paper, another promising encoding is considered, suggested by recent debates on quantum machine learning. Further, we observe a significant property concerning the non-invariance by feature rescaling of our quantum classifier. This fact, which represents a meaningful difference between the NMC and the respective quantum version, allows us to introduce a free parameter whose variation provides, in some cases, better classification results for the QNMC. The experimental section is devoted: (i) to compare the NMC and QNMC performance on different datasets; and (ii) to study the effects of the non-invariance under uniform rescaling for the QNMC.

Early fault diagnosis of bearing and stator faults of the single-phase induction motor using acoustic signals

An article describes an early fault diagnostic technique based on acoustic signals. The presented technique was used for a single-phase induction motor. The authors measured and analysed following states of the motor: healthy single-phase induction motor, single-phase induction motor with faulty bearing, single-phase induction motor with faulty bearing and shorted coils of auxiliary winding. A feature extraction method called MSAF-20-MULTIEXPANDED (Method of Selection of Amplitudes of Frequency – Multiexpanded) was discussed. The MSAF-20-MULTIEXPANDED was used to create feature vectors. The obtained vectors were classified by NN (Nearest Neighbour classifier), NM (Nearest Mean classifier) and GMM (Gaussian Mixture Models). The proposed technique can be used for diagnosis of the single-phase induction motors. It can be also used for other types of rotating electric motors.

A Quantum-inspired Version of the Classification Problem

We address the problem of binary classification by using a quantum version of the Nearest Mean Classifier (NMC). Our proposal is indeed an advanced version of previous one (see Sergioli et al. 2017 that i) is able to be naturally generalized to arbitrary number of features and ii) exhibits better performances with respect to the classical NMC for several datasets. Further, we show that the quantum version of NMC is not invariant under rescaling. This allows us to introduce a free parameter, i.e. the rescaling factor, that could be useful to get a further improvement of the classification performance.

A quantum-inspired version of the nearest mean classifier

We introduce a framework suitable for describing pattern recognition task using the mathematical language of density matrices. In particular, we provide a one-to-one correspondence between patterns and pure density operators. This correspondence enables us to: i) represent the Nearest Mean Classifier (NMC) in terms of quantum objects, ii) introduce a Quantum Classifier (QC). By comparing the QC with the NMC on different 2D datasets, we show the first classifier can provide additional information that are particularly beneficial on a classical computer with respect to the second classifier.

Simple quantum circuit for pattern recognition based on nearest mean classifier

Machine learning plays a key role in many applications such as data mining and image recognition.Classification is one subcategory under machine learning. In this paper we propose a simple quantum circuitbased on the nearest mean classifier to classified handwriting characters. Our circuit is a simplified circuit fromthe quantum support vector machine [Phys. Rev. Lett. 114, 140504 (2015)] which uses quantum matrix inversealgorithm to find optimal hyperplane that separated two different classes. In our case the hyperplane is foundusing projections and rotations on the Bloch sphere.

Quantitative heartbeat coupling measures in human-horse interaction.

We present a study focused on a quantitative estimation of a human-horse dynamic interaction. A set of measures based on magnitude and phase coupling between heartbeat dynamics of both humans and horses in three different conditions is reported: no interaction, visual/olfactory interaction and grooming. Specifically, Magnitude Squared Coherence (MSC), Mean Phase Coherence (MPC) and Dynamic Time Warping (DTW) have been used as estimators of the amount of coupling between human and horse through the analysis of their heart rate variability (HRV) time series in a group of eleven human subjects, and one horse. The rationale behind this study is that the interaction of two complex biological systems go towards a coupling process whose dynamical evolution is modulated by the kind and time duration of the interaction itself. We achieved a congruent and consistent statistical significant difference for all of the three indices. Moreover, a Nearest Mean Classifier was able to recognize the three classes of interaction with an accuracy greater than 70%. Although preliminary, these encouraging results allow a discrimination of three distinct phases in a real human-animal interaction opening to the characterization of the empirically proven relationship between human and horse.

POLLEN BEARING HONEY BEE DETECTION IN HIVE ENTRANCE VIDEO RECORDED BY REMOTE EMBEDDED SYSTEM FOR POLLINATION MONITORING

Abstract. Honey bees have crucial role in pollination across the world. This paper presents a simple, non-invasive, system for pollen bearing honey bee detection in surveillance video obtained at the entrance of a hive. The proposed system can be used as a part of a more complex system for tracking and counting of honey bees with remote pollination monitoring as a final goal. The proposed method is executed in real time on embedded systems co-located with a hive. Background subtraction, color segmentation and morphology methods are used for segmentation of honey bees. Classification in two classes, pollen bearing honey bees and honey bees that do not have pollen load, is performed using nearest mean classifier, with a simple descriptor consisting of color variance and eccentricity features. On in-house data set we achieved correct classification rate of 88.7% with 50 training images per class. We show that the obtained classification results are not far behind from the results of state-of-the-art image classification methods. That favors the proposed method, particularly having in mind that real time video transmission to remote high performance computing workstation is still an issue, and transfer of obtained parameters of pollination process is much easier.

POLLEN BEARING HONEY BEE DETECTION IN HIVE ENTRANCE VIDEO RECORDED BY REMOTE EMBEDDED SYSTEM FOR POLLINATION MONITORING

Honey bees have crucial role in pollination across the world. This paper presents a simple, non-invasive, system for pollen bearing honey bee detection in surveillance video obtained at the entrance of a hive. The proposed system can be used as a part of a more complex system for tracking and counting of honey bees with remote pollination monitoring as a final goal. The proposed method is executed in real time on embedded systems co-located with a hive. Background subtraction, color segmentation and morphology methods are used for segmentation of honey bees. Classification in two classes, pollen bearing honey bees and honey bees that do not have pollen load, is performed using nearest mean classifier, with a simple descriptor consisting of color variance and eccentricity features. On in-house data set we achieved correct classification rate of 88.7% with 50 training images per class. We show that the obtained classification results are not far behind from the results of state-of-the-art image classification methods. That favors the proposed method, particularly having in mind that real time video transmission to remote high performance computing workstation is still an issue, and transfer of obtained parameters of pollination process is much easier.

Hierarchical learning of large-margin metrics for large-scale image classification

Large-scale image classification is a challenging task and has recently attracted active research interests. In this paper, a new algorithm is developed to achieve more effective implementation of large-scale image classification by hierarchical learning of large-margin metrics (HLMMs). A hierarchical visual tree is seamlessly integrated with metric learning to learn a set of node-specific/category-specific large-margin metrics. First, a hierarchical visual tree is learned to characterize the inter-category visual correlations effectively and organize large numbers of image categories in a coarse-to-fine fashion. Second, a new algorithm is developed to support hierarchical learning of large-margin metrics by training nearest class mean (NCM) classifiers over our hierarchical visual tree. In addition, we also consider dimensionality reduction as a regularizer for high-dimensional data in our large-margin metric learning. Two top-down approaches are developed for supporting hierarchical learning of large-margin metrics. We focus on learning more discriminative metrics for NCM node classifiers to identify the visually similar sub-nodes (visually similar image categories) under the same parent node over our hierarchical visual tree. A mini-batch stochastic gradient descend method is used to optimize our HLMMs learning algorithm. The experimental results on ImageNet Large Scale Visual Recognition Challenge 2010 dataset (ILSVRC2010) have demonstrated that our HLMMs learning algorithm is very promising for supporting large-scale image classification.

Vision Based Multi-Feature Hand Gesture Recognition For Indian Sign Language Manual Signs

Abstract Indian sign language (ISL) is the main communication medium among deaf Indians. An ISL vocabulary show that the hand plays a significant role in ISL. ISL includes static and dynamic hand gesture recognition. The main aim of this paper is to present multi-feature static hand gesture recognition for alphabets and numbers. Here, comparative analysis of various feature descriptors such as chain code, shape matrix, Fourier descriptor, 7 Hu moments, and boundary moments is done. Multi-feature fusion descriptor is designed using contour (Boundary moments, Fourier descriptor) and region based (7Hu moments) descriptors. Analysis of this new multi-feature descriptor is done in comparison with other individual descriptors and it showed noteworthy results over other descriptors. Three classification methods such as, Nearest Mean Classifier (NMC), k-Nearest Neighborhood (k-NN) and Naive Bayes classifier are used for classification and comparison. New Multi-feature fusion descriptor shows high recognition rate of 99.61% among all with k-NN. Real time recognition for number signs 0-9, of fusion descriptor with NMC gave 100% accuracy

Semi-supervised nearest mean classification through a constrained log-likelihood.

We cast a semi-supervised nearest mean classifier, previously introduced by the first author, in a more principled log-likelihood formulation that is subject to constraints. This, in turn, leads us to make the important suggestion to not only investigate error rates of semi-supervised learners but also consider the risk they originally aim to optimize. We demonstrate empirically that in terms of classification error, mixed results are obtained when comparing supervised to semi-supervised nearest mean classification, while in terms of log-likelihood on the test set, the semi-supervised method consistently outperforms its supervised counterpart. Comparisons to self-learning, a standard approach in semi-supervised learning, are included to further clarify the way, in which our constrained nearest mean classifier improves over regular, supervised nearest mean classification.

Medical Data Classifications Using Genetic Algorithm Based Generalized Kernel Linear Discriminant Analysis

The generalized Kernel Linear Discriminant Analysis (KLDA) is the dimensionality reduction technique with class discrimination to map the vectors from the feature dimensional space to the lower dimensional space. In this paper, we propose to tune the unknown parameters of the generalized KLDA using genetic algorithm to map the vectors from the feature dimensional space to the lower dimensional space. Nearest mean classifier is used for classification. Experiments are performed on medical data using the genetic algorithm based GLDA and reported in this paper. As an average 5% increase in the detection rate is achieved using the genetic algorithm based GLDA when compared with the other kernel function based LDA.

Automated Diagnosis of Mammogram Images of Breast Cancer Using Discrete Wavelet Transform and Spherical Wavelet Transform Features: A Comparative Study

Mammograms are one of the most widely used techniques for preliminary screening of breast cancers. There is great demand for early detection and diagnosis of breast cancer using mammograms. Texture based feature extraction techniques are widely used for mammographic image analysis. In specific, wavelets are a popular choice for texture analysis of these images. Though discrete wavelets have been used extensively for this purpose, spherical wavelets have rarely been used for Computer-Aided Diagnosis (CAD) of breast cancer using mammograms. In this work, a comparison of the performance between the features of Discrete Wavelet Transform (DWT) and Spherical Wavelet Transform (SWT) based on the classification results of normal, benign and malignant stage was studied. Classification was performed using Linear Discriminant Classifier (LDC), Quadratic Discriminant Classifier (QDC), Nearest Mean Classifier (NMC), Support Vector Machines (SVM) and Parzen Classifier (ParzenC). We have obtained a maximum classification accuracy of 81.73% for DWT and 88.80% for SWT features using SVM classifier.

Recognition of Monochrome Thermal Images of Synchronous Motor with the Application of Binarization and Nearest Mean Classifier

Abstract This article discusses the recognition method of imminent failure conditions of synchronous motor. The proposed approach is based on a study of thermal images of the motor. Studies were carried out for four conditions of motor with the application of binarization and nearest mean classifier with Manhattan distance. Pattern creation process used 40 monochrome thermal images. Identification process was carried out for 160 monochrome thermal images. The experiments show that the method can be useful for protection of synchronous motor. Moreover, this method can be used to diagnose equipments in steelworks and other industrial plants.

One-Class Classification of Mammograms Using Trace Transform Functionals

Mammography is one of the first diagnostic tests to prescreen breast cancer. Early detection of breast cancer has been known to improve recovery rates to a great extent. In most medical centers, experienced radiologists are given the responsibility of analyzing mammograms. But, there is always a possibility of human error. Errors can frequently occur as a result of fatigue of the observer, resulting in interobserver and intraobserver variations. The sensitivity of mammographic screening also varies with image quality. To offset different kinds of variability and to standardize diagnostic procedures, efforts are being made to develop automated techniques for diagnosis and grading of breast cancer images. This paper presents a one-class classification pipeline for the classification of breast cancer images into benign and malignant classes. Because of the sparse distribution of abnormal mammograms, the two-class classification problem is reduced to a one-class outlier identification problem. Trace transform, which is a generalization of the Radon transform, has been used to extract the features. Several new functionals specific to mammographic image analysis have been developed and implemented to yield clinically significant features. Classifiers such as the linear discriminant classifier, quadratic discriminant classifier, nearest mean classifier, support vector machine, and the Gaussian mixture model (GMM) were used. For automated diagnosis, the classification pipeline was tested on a set of 313 mammograms provided by the Singapore Anti-Tuberculosis Association CommHealth. A maximum accuracy rate of 92.48% has been obtained using GMMs.

A novel steady-state visually evoked potential-based brain–computer interface design: Character Plotter

In this paper, an electroencephalogram-based innovative brain–computer interface (BCI), known as “Character Plotter”, is presented. The proposed design uses steady-state visually evoked potentials. The subjects generate characters by drawing, step by step, lines between six circles on the computer screen. Additionally, there are three circles for controlling the drawing procedure. The features obtained from canonical correlation analysis are used for classification. The proposed synchronous BCI design was tested on 16 subjects in offline and online experimental tasks using support vector machines, linear discriminant analysis and the nearest mean classifier. The obtained classification performances indicate that the proposed design can be successfully used in classification. The Character Plotter converges to the natural writing scheme of humans. Therefore, subjects can adapt to the BCI design in a short training session.

Finger Knuckle Identification using Principal Component Analysis and Nearest Mean Classifier

The texture pattern produced by the finger knuckle bending is highly unique and makes the surface a distinctive biometric identifier. This paper presents literature survey and classification method for an emerging biometric identifier, namely Finger-Knuckle-Print (FKP), for personal identification. The FKP feature extraction is done using Principal Component Analysis (PCA) technique. Also Knuckle classification using nearest mean classifier is proposed in this paper. The experimental results from the proposed approach are promising and confirm the usefulness of this approach for personal identification.

Distance-Based Image Classification: Generalizing to New Classes at Near-Zero Cost

We study large-scale image classification methods that can incorporate new classes and training images continuously over time at negligible cost. To this end, we consider two distance-based classifiers, the k-nearest neighbor (k-NN) and nearest class mean (NCM) classifiers, and introduce a new metric learning approach for the latter. We also introduce an extension of the NCM classifier to allow for richer class representations. Experiments on the ImageNet 2010 challenge dataset, which contains over 10(6) training images of 1,000 classes, show that, surprisingly, the NCM classifier compares favorably to the more flexible k-NN classifier. Moreover, the NCM performance is comparable to that of linear SVMs which obtain current state-of-the-art performance. Experimentally, we study the generalization performance to classes that were not used to learn the metrics. Using a metric learned on 1,000 classes, we show results for the ImageNet-10K dataset which contains 10,000 classes, and obtain performance that is competitive with the current state-of-the-art while being orders of magnitude faster. Furthermore, we show how a zero-shot class prior based on the ImageNet hierarchy can improve performance when few training images are available.

Decision support system for breast cancer detection using mammograms

Mammograms are by far one of the most preferred methods of screening for breast cancer. Early detection of breast cancer can improve survival rates to a greater extent. Although the analysis and diagnosis of breast cancer are done by experienced radiologists, there is always the possibility of human error. Interobserver and intraobserver errors occur frequently in the analysis of medical images, given the high variability between every patient. Also, the sensitivity of mammographic screening varies with image quality and expertise of the radiologist. So, there is no golden standard for the screening process. To offset this variability and to standardize the diagnostic procedures, efforts are being made to develop automated techniques for diagnosis and grading of breast cancer images. This article presents a classification pipeline to improve the accuracy of differentiation between normal, benign, and malignant mammograms. Several features based on higher-order spectra, local binary pattern, Laws' texture energy, and discrete wavelet transform were extracted from mammograms. Feature selection techniques based on sequential forward, backward, plus-l-takeaway-r, individual, and branch-and-bound selections using the Mahalanobis distance criterion were used to rank the features and find classification accuracies for combination of several features based on the ranking. Six classifiers were used, namely, decision tree classifier, fisher classifier, linear discriminant classifier, nearest mean classifier, Parzen classifier, and support vector machine classifier. We evaluated our proposed methodology with 300 mammograms obtained from the Digital Database for Screening Mammography and 300 mammograms from the Singapore Anti-Tuberculosis Association CommHealth database. Sensitivity, specificity, and accuracy values were used to compare the performances of the classifiers. Our results show that the decision tree classifier demonstrated an excellent performance compared to other classifiers with classification accuracy, sensitivity, and specificity of 91% for the Digital Database for Screening Mammography database and 96.8% for the Singapore Anti-Tuberculosis Association CommHealth database.

Integration of Clinical and Gene Expression Data Has a Synergetic Effect on Predicting Breast Cancer Outcome

Breast cancer outcome can be predicted using models derived from gene expression data or clinical data. Only a few studies have created a single prediction model using both gene expression and clinical data. These studies often remain inconclusive regarding an obtained improvement in prediction performance. We rigorously compare three different integration strategies (early, intermediate, and late integration) as well as classifiers employing no integration (only one data type) using five classifiers of varying complexity. We perform our analysis on a set of 295 breast cancer samples, for which gene expression data and an extensive set of clinical parameters are available as well as four breast cancer datasets containing 521 samples that we used as independent validation.mOn the 295 samples, a nearest mean classifier employing a logical OR operation (late integration) on clinical and expression classifiers significantly outperforms all other classifiers. Moreover, regardless of the integration strategy, the nearest mean classifier achieves the best performance. All five classifiers achieve their best performance when integrating clinical and expression data. Repeating the experiments using the 521 samples from the four independent validation datasets also indicated a significant performance improvement when integrating clinical and gene expression data. Whether integration also improves performances on other datasets (e.g. other tumor types) has not been investigated, but seems worthwhile pursuing. Our work suggests that future models for predicting breast cancer outcome should exploit both data types by employing a late OR or intermediate integration strategy based on nearest mean classifiers.