Particle Competition and Cooperation in Networks for Semi-Supervised Learning with Concept Drift

Concept drift is a problem of increasing importance in machine learning and data mining. Data sets under analysis are no longer only static databases, but also data streams in which concepts and data distributions may not be stable over time. However, most learning algorithms produced so far are based on the assumption that data comes from a fixed distribution, so they are not suitable to handle concept drifts. Moreover, some concept drifts applications requires fast response, which means an algorithm must always be (re) trained with the latest available data. But the process of labeling data is usually expensive and/or time consuming when compared to unlabeled data acquisition, thus only a small fraction of the incoming data may be effectively labeled. Semi-supervised learning methods may help in this scenario, as they use both labeled and unlabeled data in the training process. However, most of them are also based on the assumption that the data is static. Therefore, semi-supervised learning with concept drifts is still an open challenge in machine learning. Recently, a particle competition and cooperation approach was used to realize graph-based semi-supervised learning from static data. In this paper, we extend that approach to handle data streams and concept drift. The result is a passive algorithm using a single classifier, which naturally adapts to concept changes, without any explicit drift detection mechanism. Its built-in mechanisms provide a natural way of learning from new data, gradually forgetting older knowledge as older labeled data items became less influent on the classification of newer data items. Some computer simulation are presented, showing the effectiveness of the proposed method.

Context-sensitive analysis of obfuscated x86 executables

A method for context-sensitive analysis of binaries that may have obfuscated procedure call and return operations is presented. Such binaries may use operators to directly manipulate stack instead of using native call and ret instructions to achieve equivalent behavior. Since definition of context-sensitivity and algorithms for context-sensitive analysis have thus far been based on the specific semantics associated to procedure call and return operations, classic interprocedural analyses cannot be used reliably for analyzing programs in which these operations cannot be discerned. A new notion of context-sensitivity is introduced that is based on the state of the stack at any instruction. While changes in 'calling'-context are associated with transfer of control, and hence can be reasoned in terms of paths in an interprocedural control flow graph (ICFG), the same is not true of changes in 'stack'-context. An abstract interpretation based framework is developed to reason about stack-contexts and to derive analogues of call-strings based methods for the context-sensitive analysis using stack-context. The method presented is used to create a context-sensitive version of Venable et al.'s algorithm for detecting obfuscated calls. Experimental results show that the context-sensitive version of the algorithm generates more precise results and is also computationally more efficient than its context-insensitive counterpart. Copyright © 2010 ACM.

A PZT-based technique for SHM using the coherence function

This paper presents a new approach for damage detection in structural health monitoring systems exploiting the coherence function between the signals from PZT (Lead Zirconate Titanate) transducers bonded to a host structure. The physical configuration of this new approach is similar to the configuration used in Lamb wave based methods, but the analysis and operation are different. A PZT excited by a signal with a wide frequency range acts as an actuator and others PZTs are used as sensors to receive the signal. The coherences between the signals from the PZT sensors are obtained and the standard deviation for each coherence function is computed. It is demonstrated through experimental results that the standard deviation of the coherence between the signals from the PZTs in healthy and damaged conditions is a very sensitive metric index to detect damage. Tests were carried out on an aluminum plate and the results show that the proposed methodology could be an excellent approach for structural health monitoring (SHM) applications.

Particle competition and cooperation to prevent error propagation from mislabeled data in semi-supervised learning

Semi-supervised learning is applied to classification problems where only a small portion of the data items is labeled. In these cases, the reliability of the labels is a crucial factor, because mislabeled items may propagate wrong labels to a large portion or even the entire data set. This paper aims to address this problem by presenting a graph-based (network-based) semi-supervised learning method, specifically designed to handle data sets with mislabeled samples. The method uses teams of walking particles, with competitive and cooperative behavior, for label propagation in the network constructed from the input data set. The proposed model is nature-inspired and it incorporates some features to make it robust to a considerable amount of mislabeled data items. Computer simulations show the performance of the method in the presence of different percentage of mislabeled data, in networks of different sizes and average node degree. Importantly, these simulations reveals the existence of the critical points of the mislabeled subset size, below which the network is free of wrong label contamination, but above which the mislabeled samples start to propagate their labels to the rest of the network. Moreover, numerical comparisons have been made among the proposed method and other representative graph-based semi-supervised learning methods using both artificial and real-world data sets. Interestingly, the proposed method has increasing better performance than the others as the percentage of mislabeled samples is getting larger. © 2012 IEEE.

Traditional and novel methods for occlusal caries detection: Performance on primary teeth

This study aimed to assess the performance of International Caries Detection and Assessment System (ICDAS), radiographic examination, and fluorescence-based methods for detecting occlusal caries in primary teeth. One occlusal site on each of 79 primary molars was assessed twice by two examiners using ICDAS, bitewing radiography (BW), DIAGNOdent 2095 (LF), DIAGNOdent 2190 (LFpen), and VistaProof fluorescence camera (FC). The teeth were histologically prepared and assessed for caries extent. Optimal cutoff limits were calculated for LF, LFpen, and FC. At the D 1 threshold (enamel and dentin lesions), ICDAS and FC presented higher sensitivity values (0.75 and 0.73, respectively), while BW showed higher specificity (1.00). At the D 2 threshold (inner enamel and dentin lesions), ICDAS presented higher sensitivity (0.83) and statistically significantly lower specificity (0.70). At the D 3 threshold (dentin lesions), LFpen and FC showed higher sensitivity (1.00 and 0.91, respectively), while higher specificity was presented by FC (0.95), ICDAS (0.94), BW (0.94), and LF (0.92). The area under the receiver operating characteristic (ROC) curve (Az) varied from 0.780 (BW) to 0.941 (LF). Spearman correlation coefficients with histology were 0.72 (ICDAS), 0.64 (BW), 0.71 (LF), 0.65 (LFpen), and 0.74 (FC). Inter- and intraexaminer intraclass correlation values varied from 0.772 to 0.963 and unweighted kappa values ranged from 0.462 to 0.750. In conclusion, ICDAS and FC exhibited better accuracy in detecting enamel and dentin caries lesions, whereas ICDAS, LF, LFpen, and FC were more appropriate for detecting dentin lesions on occlusal surfaces in primary teeth, with no statistically significant difference among them. All methods presented good to excellent reproducibility. © 2012 Springer-Verlag London Ltd.

ECG arrhythmia classification based on optimum-path forest

An important tool for the heart disease diagnosis is the analysis of electrocardiogram (ECG) signals, since the non-invasive nature and simplicity of the ECG exam. According to the application, ECG data analysis consists of steps such as preprocessing, segmentation, feature extraction and classification aiming to detect cardiac arrhythmias (i.e.; cardiac rhythm abnormalities). Aiming to made a fast and accurate cardiac arrhythmia signal classification process, we apply and analyze a recent and robust supervised graph-based pattern recognition technique, the optimum-path forest (OPF) classifier. To the best of our knowledge, it is the first time that OPF classifier is used to the ECG heartbeat signal classification task. We then compare the performance (in terms of training and testing time, accuracy, specificity, and sensitivity) of the OPF classifier to the ones of other three well-known expert system classifiers, i.e.; support vector machine (SVM), Bayesian and multilayer artificial neural network (MLP), using features extracted from six main approaches considered in literature for ECG arrhythmia analysis. In our experiments, we use the MIT-BIH Arrhythmia Database and the evaluation protocol recommended by The Association for the Advancement of Medical Instrumentation. A discussion on the obtained results shows that OPF classifier presents a robust performance, i.e.; there is no need for parameter setup, as well as a high accuracy at an extremely low computational cost. Moreover, in average, the OPF classifier yielded greater performance than the MLP and SVM classifiers in terms of classification time and accuracy, and to produce quite similar performance to the Bayesian classifier, showing to be a promising technique for ECG signal analysis. © 2012 Elsevier Ltd. All rights reserved.

Context-sensitive analysis of x86 obfuscated executables

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Semi-supervised learning with concept drift using particle dynamics applied to network intrusion detection data

Concept drift, which refers to non stationary learning problems over time, has increasing importance in machine learning and data mining. Many concept drift applications require fast response, which means an algorithm must always be (re)trained with the latest available data. But the process of data labeling is usually expensive and/or time consuming when compared to acquisition of unlabeled data, thus usually only a small fraction of the incoming data may be effectively labeled. Semi-supervised learning methods may help in this scenario, as they use both labeled and unlabeled data in the training process. However, most of them are based on assumptions that the data is static. Therefore, semi-supervised learning with concept drifts is still an open challenging task in machine learning. Recently, a particle competition and cooperation approach has been developed to realize graph-based semi-supervised learning from static data. We have extend that approach to handle data streams and concept drift. The result is a passive algorithm which uses a single classifier approach, naturally adapted to concept changes without any explicit drift detection mechanism. It has built-in mechanisms that provide a natural way of learning from new data, gradually "forgetting" older knowledge as older data items are no longer useful for the classification of newer data items. The proposed algorithm is applied to the KDD Cup 1999 Data of network intrusion, showing its effectiveness.

Particle competition and cooperation for semi-supervised learning with label noise

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Doxycycline hyclate: a review of properties, applications and analytical methods

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Transmission system expansion planning by an extended genetic algorithm

The paper presents an extended genetic algorithm for solving the optimal transmission network expansion planning problem. Two main improvements have been introduced in the genetic algorithm: (a) initial population obtained by conventional optimisation based methods; (b) mutation approach inspired in the simulated annealing technique, the proposed method is general in the sense that it does not assume any particular property of the problem being solved, such as linearity or convexity. Excellent performance is reported in the test results section of the paper for a difficult large-scale real-life problem: a substantial reduction in investment costs has been obtained with regard to previous solutions obtained via conventional optimisation methods and simulated annealing algorithms; statistical comparison procedures have been employed in benchmarking different versions of the genetic algorithm and simulated annealing methods.

Genotyping of AR and PSA polymorphisms in a patient with Klinefelter syndrome, non-Hodgkin lymphoma, and adenocarcinoma of the prostate

It has been hypothesized that the AR (androgen receptor) gene binds the two PSA (prostate-specific antigen) alleles with differing affinities and may differentially influence prostate cancer risk. In this article, we report a case of adenocarcinoma of the prostate in a 56-year-old man with Klinefelter syndrome (47,XXY) and non-Hodgkin lymphoma, as well as the AR and PSA genotype. AR and PSA gene polymorphisms were analyzed by polymerase chain reaction-based methods using DNA from peripheral white blood cells and the prostate cancer. We determined the methylation status of the AR gene on the X chromosome. The patient presents with the AG genotype for the ARE-I (androgen response element) region of the PSA gene. We detect the presence of two short AR alleles with 19 and 11CAG repeats each. Unmethylated alleles were demonstrated for both. The shorter allele was inactive in more than 60% of total DNA in both control blood and prostate cancer cells. The presence of short AR alleles and the G allele of the PSA gene may contribute to the development of prostate cancer in a 47,XXY patient. (C) 2004 Elsevier B.V. All rights reserved.

Detection of Babesia bigemina in cattle of different genetic groups and in Rhipicephalus (Boophilus) microplus tick

Babesia bigemina infections were investigated in four genetic groups of beef cattle and in Rhipicephalus (Boophilus) microplus engorged female ticks. Blood samples and engorged female ticks were collected from 15 cows and 15 calves from each of the following genetic groups: Nelore, Angus x Nelore, Canchim x Nelore, and Simmental x Nelore. Microscopic examination of blood smears and tick hemolymph revealed that merozoites of B. bigemina (6/60) as well as kinetes of Babesia spp. (9/549) were only detected in samples (blood and ticks, respectively) originated from calves. PCR-based methods using primers for specific detection of B. bigemina revealed 100% infection in both calves and cows, regardless the genetic group. Tick infection was detected by nested-PCR amplifications showing that the frequency of B. bigemina was higher (P 0.01) in female ticks collected from calves (134/549) than in those collected from cows (52/553). The frequency of B. bigemina was similar in ticks collected from animals, either cows or calves, of the four genetic groups (P > 0.05). (C) 2008 Elsevier B.V. All rights reserved.

Anchor Areas to Improve Conservation and Increase Connectivity within the Brazilian "Mesopotamia of Biodiversity"

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Delimiting genetic units in Neotropical toads under incomplete lineage sorting and hybridization

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)