基于可拒识-双层支持向量分类器的微钙化点检测

为克服医学图像微钙化点检测中假阳性高的缺点,构造了一种带拒识能力的双层支持向量模型分类器,用于钙化点检测.检测时,首先利用基于最大间隔超平面的支持向量分类器（SVC）对输入模式进行分类判决;然后通过求取真实钙化点样本特征空间最小的包含球形边界来得到钙化点样本的球形支持向量域表示（SVDD）;接着利用钙化点的支持向量域表示对输入模式进行拒识或接受处理;最后利用SVC与SVDD两个分类器的结果来进行综合判决.仿真实验结果表明,该算法在不影响微钙化点的检出率的情况下,可部分解决假阳性高的问题.     

Comparative study of supervised classification algorithms for the detection of atmospheric pollution

The management of atmospheric pollution using video is not yet widespread. However it is an efficient way to evaluate the polluting rejects coming from large industrial facilities when traditional captors are not usable. This paper presents a comparison of different classifiers for a monitoring system of polluting smokes. The data used in this work are stemming from a system of video analysis and signal processing. The database includes the pollution level of puffs of smoke defined by an expert. Six machine learning techniques are tested and compared to classify the puffs of smoke: k-nearest neighbour, naïve Bayes classifier, artificial neural network, decision tree, support vector machine and a fuzzy model. The parameters of each type of classifier are split into three categories: learned parameters, parameters determined by a first step of the experimentation, and parameters set by the programmer. We compare the results of the best classifier of each type depending on the size of the learning set. A part of the discussion concerns the robustness of the classifier facing the case where classes of interest are under-represented, as the high level of pollution in our data.     

Automatic approval prediction for software enhancement requests

Software applications often receive a large number of enhancement requests that suggest developers to fulfill additional functions. Such requests are usually checked manually by the developers, which is time consuming and tedious. Consequently, an approach that can automatically predict whether a new enhancement report will be approved is beneficial for both the developers and enhancement suggesters. With the approach, according to their available time, the developers can rank the reports and thus limit the number of reports to evaluate from large collection of low quality enhancement requests that are unlikely to be approved. The approach can help developers respond to the useful requests more quickly. To this end, we propose a multinomial naive Bayes based approach to automatically predict whether a new enhancement report is likely to be approved or rejected. We acquire the enhancement reports of open-source software applications from Bugzilla for evaluation. Each report is preprocessed and modeled as a vector. Using these vectors with their corresponding approval status, we train a Bayes based classifier. The trained classifier predicts approval or rejection of the new enhancement reports. We apply different machine learning and neural network algorithms, and it turns out that the multinomial naive Bayes classifier yields the highest accuracy with the given dataset. The proposed approach is evaluated with 40,000 enhancement reports from 35 open source applications. The results of tenfold cross validation suggest that the average accuracy is up to 89.25%.     

Large margin mixture of AR models for time series classification

In this paper, we propose the large margin autoregressive (LMAR) model for classification of time series patterns. The parameters of the generative AR models for different classes are estimated using the margin of the boundaries of AR models as the optimization criterion. Models that use a mixture of AR (MAR) models are considered for representing the data that cannot be adequately represented using a single AR model for a class. Based on a mixture model representing each class, we propose the large margin mixture of AR (LMMAR) models. The proposed methods are applied on the simulated time series data, electrocardiogram data, speech data for E-set in English alphabet and electroencephalogram time series data. Performance of the proposed methods is compared with that of support vector machine (SVM) based classifier that uses AR coefficients based features. The proposed methods give a better classification performance compared to the SVM based classifier. Being generative models, the LMAR and LMMAR models provide a generative interpretation that enables utilization of the rejection option in the high risk classification tasks. The proposed methods can also be used for detection of novel time series data.     

A novel method for automatic modulation recognition

Automatic recognition of the digital modulation plays an important role in various applications. This paper investigates the design of an accurate system for recognition of digital modulations. First, it is introduced an efficient pattern recognition system that includes two main modules: the feature extraction module and the classifier module. Feature extraction module extracts a suitable combination of the higher order moments up to eighth, higher order cumulants up to eighth and instantaneous characteristics of digital modulations. These combinations of the features are applied for the first time in this area. In the classifier module, two important classes of supervised classifiers, i.e., multi-layer perceptron (MLP) neural network and hierarchical multi-class support vector machine based classifier are investigated. By experimental study, we choose the best classifier for recognition of the considered modulations. Then, we propose a hybrid heuristic recognition system that an optimization module is added to improve the generalization performance of the classifier. In this module we have used a new optimization algorithm called Bees Algorithm. This module optimizes the classifier design by searching for the best value of the parameters that tune its discriminant function, and upstream by looking for the best subset of features that feed the classifier. Simulation results show that the proposed hybrid intelligent technique has very high recognition accuracy even at low levels of SNR with a little number of the features.     

Optimum tradeoff between class-selective rejection error and average number of classes

The paper first reviews the recently proposed optimum class-selective rejection rule. This rule provides an optimum tradeoff between the error rate and the average number of (selected) classes. Then, a new general relation between the error rate and the average number of classes is presented. The error rate can be directly computed from the class-selective reject function, which in turn can be estimated from unlabelled patterns, by simply counting the rejected classes. Theoretical as well as practical implications are discussed, and some future research directions are proposed.     

Extraction and optimization of B-spline PBD templates forrecognition of connected handwritten digit strings

The recognition of connected handwritten digit strings is a challenging task due mainly to two problems: poor character segmentation and unreliable isolated character recognition. The authors first present a rational B-spline representation of digit templates based on Pixel-to-Boundary Distance (PBD) maps. We then present a neural network approach to extract B-spline PBD templates and an evolutionary algorithm to optimize these templates. In total, 1000 templates (100 templates for each of 10 classes) were extracted from and optimized on 10426 training samples from the NIST Special Database 3. By using these templates, a nearest neighbor classifier can successfully reject 90.7 percent of nondigit patterns while achieving a 96.4 percent correct classification of isolated test digits. When our classifier is applied to the recognition of 4958 connected handwritten digit strings (4555 2-digit, 355 3-digit, and 48 4-digit strings) from the NIST Special Database 3 with a dynamic programming approach, it has a correct classification rate of 82.4 percent with a rejection rate of as low as 0.85 percent. Our classifier compares favorably in terms of correct classification rate and robustness with other classifiers that are tested     

Fingerprint classification using a feedback-based line detector

We present a fingerprint classification algorithm in this paper. This algorithm classifies a fingerprint image into one of the five classes: Arch, Left loop, Right loop, Whorl, and Tented arch. We use a new low-dimensional feature vector obtained from the output of a novel oriented line detector presented here. Our line detector is a co-operative dynamical system that gives oriented lines and preserves multiple orientations at points where differently oriented lines meet. Our feature extraction process is based on characterizing the distribution of orientations around the fingerprint. We discuss three different classifiers: support vector machines, nearest-neighbor classifier, and neural network classifier. We present results obtained on a National Institute of Standards and Technology (NIST) fingerprint database and compare with other published results on NIST databases. All our classifiers perform equally well, and this suggests that our novel line detection and feature extraction process indeed captures all the crucial information needed for classification in this problem.     

A hybrid neural network classifier combining ordered fuzzy ARTMAP and the dynamic decay adjustment algorithm

This paper presents a novel conflict-resolving neural network classifier that combines the ordering algorithm, fuzzy ARTMAP (FAM), and the dynamic decay adjustment (DDA) algorithm, into a unified framework. The hybrid classifier, known as Ordered FAMDDA, applies the DDA algorithm to overcome the limitations of FAM and ordered FAM in achieving a good generalization/performance. Prior to network learning, the ordering algorithm is first used to identify a fixed order of training patterns. The main aim is to reduce and/or avoid the formation of overlapping prototypes of different classes in FAM during learning. However, the effectiveness of the ordering algorithm in resolving overlapping prototypes of different classes is compromised when dealing with complex datasets. Ordered FAMDDA not only is able to determine a fixed order of training patterns for yielding good generalization, but also is able to reduce/resolve overlapping regions of different classes in the feature space for minimizing misclassification during the network learning phase. To illustrate the effectiveness of Ordered FAMDDA, a total of ten benchmark datasets are experimented. The results are analyzed and compared with those from FAM and Ordered FAM. The outcomes demonstrate that Ordered FAMDDA, in general, outperforms FAM and Ordered FAM in tackling pattern classification problems.     

A combination of statistical and syntactical pattern recognition applied to classification of unconstrained handwritten numerals

This paper describes a combination of statistical and syntactical pattern recognition methods. The hierarchically structured recognition system consists of a conventional statistical classifier, a structural classifier analysing the topological composition of the patterns, a stage reducing the number of hypotheses made by the first two stages, and a mixed stage based on a search for maximum similarity between syntactically generated prototypes and patterns. The stages work on different principles to avoid mistakes made in one stage by the others. This concept is applied to the recognition of numerals written without constraints. If no samples are rejected a recognition rate of 99.5% is obtained.     

基于SVM的三阶段人脸检测方法的研究与应用

为了提高人脸的检测速度,描述了一种基于SVM的三阶段人脸检测方法,加速了复杂背景下人脸的检测。在该方法中,通过前两个阶段的瀑布型分类器将大量的简单非人脸模块快速过滤掉,并在最后的第三个阶段采用了非线性SVM分类器来更高效地区分24*24像素的人脸或者非人脸图像。     

An evaluation of the neocognitron

We describe a sequence of experiments investigating the strengths and limitations of Fukushima's neocognitron as a handwritten digit classifier. Using the results of these experiments as a foundation, we propose and evaluate improvements to Fukushima's original network in an effort to obtain higher recognition performance. The neocognitron performance is shown to be strongly dependent on the choice of selectivity parameters and we present two methods to adjust these variables. Performance of the network under the more effective of the two new selectivity adjustment techniques suggests that the network fails to exploit the features that distinguish different classes of input data. To avoid this shortcoming, the network's final layer cells were replaced by a nonlinear classifier (a multilayer perceptron) to create a hybrid architecture. Tests of Fukushima's original system and the novel systems proposed in this paper suggest that it may be difficult for the neocognitron to achieve the performance of existing digit classifiers due to its reliance upon the supervisor's choice of selectivity parameters and training data.     

A multichannel approach to fingerprint classification

Fingerprint classification provides an important indexing mechanism in a fingerprint database. An accurate and consistent classification can greatly reduce fingerprint matching time for a large database. We present a fingerprint classification algorithm which is able to achieve an accuracy better than previously reported in the literature. We classify fingerprints into five categories: whorl, right loop, left loop, arch, and tented arch. The algorithm uses a novel representation (FingerCode) and is based on a two-stage classifier to make a classification. It has been tested on 4000 images in the NIST-4 database. For the five-class problem, a classification accuracy of 90 percent is achieved (with a 1.8 percent rejection during the feature extraction phase). For the four-class problem (arch and tented arch combined into one class), we are able to achieve a classification accuracy of 94.8 percent (with 1.8 percent rejection). By incorporating a reject option at the classifier, the classification accuracy can be increased to 96 percent for the five-class classification task, and to 97.8 percent for the four-class classification task after a total of 32.5 percent of the images are rejected     

Unsupervised Mode of Rejection of Foreign Patterns

The study deals with an issue of recognition of native (proper) patterns and rejection of foreign (erroneous) patterns. We present a novel unsupervised approach to rejecting foreign patterns. We construct a geometrical model, which identifies regions in the feature space that are predominantly occupied by native patterns and determines regions where foreign patterns are localized. The model is constructed in an unsupervised mode: we engage clustering to discover structures in the data and use the revealed geometry to form regions with high likelihood of being occupied by native patterns and regions in which foreign patterns are likely to be localized. The geometry of the region of rejected patterns is adjusted by two parameters, which are tuned to achieve a sound balance between rejection of foreign patterns and acceptance of native patterns. It is shown that the proposed method is applicable not only to multiclass data processing problems, but it could also be beneficial in situations when the only available information concerns a single phenomenon (a so-called a one-class data). We demonstrate the usefulness of the proposed approach by studying several publicly available medical datasets.     

Edit distance-based kernel functions for structural pattern classification

A common approach in structural pattern classification is to define a dissimilarity measure on patterns and apply a distance-based nearest-neighbor classifier. In this paper, we introduce an alternative method for classification using kernel functions based on edit distance. The proposed approach is applicable to both string and graph representations of patterns. By means of the kernel functions introduced in this paper, string and graph classification can be performed in an implicit vector space using powerful statistical algorithms. The validity of the kernel method cannot be established for edit distance in general. However, by evaluating theoretical criteria we show that the kernel functions are nevertheless suitable for classification, and experiments on various string and graph datasets clearly demonstrate that nearest-neighbor classifiers can be outperformed by support vector machines using the proposed kernel functions.     

Blind digital modulation classification in software radio using the optimized classifier and feature subset selection

Automatic recognition of digital modulations plays an important role in various applications such as software defined radio. This study investigates the design of an accurate system for recognition of digital modulations. First, an efficient system is introduced that includes two main modules: the feature extraction module and the classifier module. First module extracts a suitable combination of the higher order moments up to eighth, higher order cumulants up to eighth and instantaneous characteristics of digital modulations. These features are applied for the first time in this area. In the classifier module, several supervised classifiers, such as multilayer perceptron neural network, radial basis function and multi-class support vector machine based classifier are investigated. By experimental study, we choose the best classifier for recognition of the considered modulations. Then, we propose a hybrid heuristic recognition system to which an optimization module is added to improve the generalization performance of the classifier. This module optimizes the classifier design by searching for the best value of the parameters that tune its discriminant function (kernel parameters selection) and upstream by looking for the best subset of features that feed the classifier. Simulation results show that the proposed system has a very high recognition accuracy. This high efficiency is achieved with little features, which have been selected using particle swarm optimizer.     

Nearest-Neighbor Automatic Sound Annotation with a WordNet Taxonomy

Sound engineers need to access vast collections of sound effects for their film and video productions. Sound effects providers rely on text-retrieval techniques to give access to their collections. Currently, audio content is annotated manually, which is an arduous task. Automatic annotation methods, normally fine-tuned to reduced domains such as musical instruments or limited sound effects taxonomies, are not mature enough for labeling with great detail any possible sound. A general sound recognition tool would require first, a taxonomy that represents the world and, second, thousands of classifiers, each specialized in distinguishing little details. We report experimental results on a general sound annotator. To tackle the taxonomy definition problem we use WordNet, a semantic network that organizes real world knowledge. In order to overcome the need of a huge number of classifiers to distinguish many different sound classes, we use a nearest-neighbor classifier with a database of isolated sounds unambiguously linked to WordNet concepts. A 30% concept prediction is achieved on a database of over 50,000 sounds and over 1600 concepts.Part of the contents of this paper has been published in the Proceedings of the 2004 IEEE International Workshop on Machine Learning for Signal Processing.     

Recognition and pose estimation of unoccluded three-dimensionalobjects from a two-dimensional perspective view by banks of neuralnetworks

This paper describes a neural network (NN) based system for recognition and pose estimation of an unoccluded three-dimensional (3-D) object from any single two-dimensional (2-D) perspective view. The approach is invariant to translation, orientation, and scale. First, the binary silhouette of the object is obtained and normalized for translation and scale. Then, the object is represented by a set of rotation invariant features derived from the complex orthogonal pseudo-Zernike moments of the image. The recognition scheme combines the decisions of a bank of multilayer perceptron NN classifiers operating in parallel on the same data. These classifiers have different topologies and internal parameters, but are trained on the same set of exemplar perspective views of the objects. Next, two pose parameters, elevation and aspect angles, are obtained by a novel two-stage NN system consisting of a quadrant classifier followed by NN angle estimators. Performance is tested on clean and noisy data bases of military ground vehicles. Comparative studies with three other classifiers (a single NN, the weighted nearest-neighbor classifier, and a binary decision tree) are carried out.     

Optical character recognition of the Orthodox Hellenic Byzantine Music notation

In this paper we present for the first time, the development of a new system for the off-line optical recognition of the characters used in the orthodox Hellenic Byzantine Music notation, that has been established since 1814. We describe the structure of the new system and propose algorithms for the recognition of the 71 distinct character classes, based on Wavelets, 4-projections and other structural and statistical features. Using a nearest neighbor classifier, combined with a post classification schema and a tree-structured classification philosophy, an accuracy of 99.4% was achieved, in a database of about 18,000 Byzantine character patterns that have been developed for the needs of the system.