Information Fusion and Hierarchical Knowledge Discovery by ARTMAP Neural Networks

Mapping novel terrain from sparse, complex data often requires the resolution of conflicting information from sensors working at different times, locations, and scales, and from experts with different goals and situations. Information fusion methods help resolve inconsistencies in order to distinguish correct from incorrect answers, as when evidence variously suggests that an object's class is car, truck, or airplane. The methods developed here consider a complementary problem, supposing that information from sensors and experts is reliable though inconsistent, as when evidence suggests that an objects class is car, vehicle, or man-made. Underlying relationships among objects are assumed to be unknown to the automated system of the human user. The ARTMAP information fusion system uses distributed code representations that exploit the neural network's capacity for one-to-many learning in order to produce self-organizing expert systems that discover hierarchial knowledge structures. The system infers multi-level relationships among groups of output classes, without any supervised labeling of these relationships. The procedure is illustrated with two image examples.

Self-Organizing Hierarchical Knowledge Discovery by an ARTMAP Image Fusion System

Classifying novel terrain or objects front sparse, complex data may require the resolution of conflicting information from sensors working at different times, locations, and scales, and from sources with different goals and situations. Information fusion methods can help resolve inconsistencies, as when evidence variously suggests that an object's class is car, truck, or airplane. The methods described here consider a complementary problem, supposing that information from sensors and experts is reliable though inconsistent, as when evidence suggests that an object's class is car, vehicle, and man-made. Underlying relationships among objects are assumed to be unknown to the automated system or the human user. The ARTMAP information fusion system used distributed code representations that exploit the neural network's capacity for one-to-many learning in order to produce self-organizing expert systems that discover hierarchical knowledge structures. The system infers multi-level relationships among groups of output classes, without any supervised labeling of these relationships.

Self-Organizing Information Fusion and Hierarchical Knowledge Discovery: A New Framework Using Artmap Neural Networks

Classifying novel terrain or objects from sparse, complex data may require the resolution of conflicting information from sensors woring at different times, locations, and scales, and from sources with different goals and situations. Information fusion methods can help resolve inconsistencies, as when eveidence variously suggests that and object's class is car, truck, or airplane. The methods described her address a complementary problem, supposing that information from sensors and experts is reliable though inconsistent, as when evidence suggests that an object's class is car, vehicle, and man-made. Underlying relationships among classes are assumed to be unknown to the autonomated system or the human user. The ARTMAP information fusion system uses distributed code representations that exploit the neural network's capacity for one-to-many learning in order to produce self-organizing expert systems that discover hierachical knowlege structures. The fusion system infers multi-level relationships among groups of output classes, without any supervised labeling of these relationships. The procedure is illustrated with two image examples, but is not limited to image domain.

Temporal pattern recognition in video clips detection

Temporal representation and reasoning plays an important role in Data Mining and Knowledge Discovery, particularly, in mining and recognizing patterns with rich temporal information. Based on a formal characterization of time-series and state-sequences, this paper presents the computational technique and algorithm for matching state-based temporal patterns. As a case study of real-life applications, zone-defense pattern recognition in basketball games is specially examined as an illustrating example. Experimental results demonstrate that it provides a formal and comprehensive temporal ontology for research and applications in video events detection.

Host-based P2P flow identification and use in real-time

Data identification is a key task for any Internet Service Provider (ISP) or network administrator. As port fluctuation and encryption become more common in P2P traffic wishing to avoid identification, new strategies must be developed to detect and classify such flows. This paper introduces a new method of separating P2P and standard web traffic that can be applied as part of a data mining process, based on the activity of the hosts on the network. Unlike other research, our method is aimed at classifying individual flows rather than just identifying P2P hosts or ports. Heuristics are analysed and a classification system proposed. The accuracy of the system is then tested using real network traffic from a core internet router showing over 99% accuracy in some cases. We expand on this proposed strategy to investigate its application to real-time, early classification problems. New proposals are made and the results of real-time experiments compared to those obtained in the data mining research. To the best of our knowledge this is the first research to use host based flow identification to determine a flows application within the early stages of the connection.

Assessment of the benefits of Discrete Conditional Survival Models in modelling ambulance response times

Many of the challenges faced in health care delivery can be informed through building models. In particular, Discrete Conditional Survival (DCS) models, recently under development, can provide policymakers with a flexible tool to assess time-to-event data. The DCS model is capable of modelling the survival curve based on various underlying distribution types and is capable of clustering or grouping observations (based on other covariate information) external to the distribution fits. The flexibility of the model comes through the choice of data mining techniques that are available in ascertaining the different subsets and also in the choice of distribution types available in modelling these informed subsets. This paper presents an illustrated example of the Discrete Conditional Survival model being deployed to represent ambulance response-times by a fully parameterised model. This model is contrasted against use of a parametric accelerated failure-time model, illustrating the strength and usefulness of Discrete Conditional Survival models.

Discrete Conditional Phase-type model (DC_Ph) for patient waiting time with a logistic regression component to predict patient admission to hospital

Discrete Conditional Phase-type (DC-Ph) models are a family of models which represent skewed survival data conditioned on specific inter-related discrete variables. The survival data is modeled using a Coxian phase-type distribution which is associated with the inter-related variables using a range of possible data mining approaches such as Bayesian networks (BNs), the Naïve Bayes Classification method and classification regression trees. This paper utilizes the Discrete Conditional Phase-type model (DC-Ph) to explore the modeling of patient waiting times in an Accident and Emergency Department of a UK hospital. The resulting DC-Ph model takes on the form of the Coxian phase-type distribution conditioned on the outcome of a logistic regression model.

Exploring the Transcriptome of Atlantic Salmon (Salmo salar) Skin, a Major Defense Organ

The skin of fish is the first line of defense against pathogens and parasites. The skin transcriptome of the Atlantic salmon is poorly characterized, and currently only 2,089 expressed sequence tags (ESTs) out of a total of half a million sequences are generated from skin-derived cDNA libraries. The primary aim of this study was to enhance the transcriptomic knowledge of salmon skin by using next-generation sequencing (NGS) technology, namely the Roche-454 platform. An equimolar mixture of high-quality RNA from skin and epidermal samples of salmon reared in either freshwater or seawater was used for 454-sequencing. This technique yielded over 600,000 reads, which were assembled into 34,696 isotigs using Newbler. Of these isotigs, 12 % had not been sequenced in Atlantic salmon, hence representing previously unreported salmon mRNAs that can potentially be skin-specific. Many full-length genes have been acquired, representing numerous biological processes. Mucin proteins are the main structural component of mucus and we examined in greater detail the sequences we obtained for these genes. Several isotigs exhibited homology to mammalian mucins (MUC2, MUC5AC and MUC5B). Mucin mRNAs are generally > 10 kbp and contain large repetitive units, which pose a challenge towards full-length sequence discovery. To date, we have not unearthed any full-length salmon mucin genes with this dataset, but have both N- and C-terminal regions of a mucin type 5. This highlights the fact that, while NGS is indeed a formidable tool for sequence data mining of non-model species, it must be complemented with additional experimental and bioinformatic work to characterize some mRNA sequences with complex features.

Combined genomic and phenotype screening reveals secretory factor SPINK1 as an invasion and survival factor associated with patient prognosis in breast cancer

Secretory factors that drive cancer progression are attractive immunotherapeutic targets. We used a whole-genome data-mining approach on multiple cohorts of breast tumours annotated for clinical outcomes to discover such factors. We identified Serine protease inhibitor Kazal-type 1 (SPINK1) to be associated with poor survival in estrogen receptor-positive (ER+) cases. Immunohistochemistry showed that SPINK1 was absent in normal breast, present in early and advanced tumours, and its expression correlated with poor survival in ER+ tumours. In ER- cases, the prognostic effect did not reach statistical significance. Forced expression and/or exposure to recombinant SPINK1 induced invasiveness without affecting cell proliferation. However, down-regulation of SPINK1 resulted in cell death. Further, SPINK1 overexpressing cells were resistant to drug-induced apoptosis due to reduced caspase-3 levels and high expression of Bcl2 and phospho-Bcl2 proteins. Intriguingly, these anti-apoptotic effects of SPINK1 were abrogated by mutations of its protease inhibition domain. Thus, SPINK1 affects multiple aggressive properties in breast cancer: survival, invasiveness and chemoresistance. Because SPINK1 effects are abrogated by neutralizing antibodies, we suggest that SPINK1 is a viable potential therapeutic target in breast cancer.

EEG decoding of semantic category reveals distributed representations for single concepts

Achieving a clearer picture of categorial distinctions in the brain is essential for our understanding of the conceptual lexicon, but much more fine-grained investigations are required in order for this evidence to contribute to lexical research. Here we present a collection of advanced data-mining techniques that allows the category of individual concepts to be decoded from single trials of EEG data. Neural activity was recorded while participants silently named images of mammals and tools, and category could be detected in single trials with an accuracy well above chance, both when considering data from single participants, and when group-training across participants. By aggregating across all trials, single concepts could be correctly assigned to their category with an accuracy of 98%. The pattern of classifications made by the algorithm confirmed that the neural patterns identified are due to conceptual category, and not any of a series of processing-related confounds. The time intervals, frequency bands and scalp locations that proved most informative for prediction permit physiological interpretation: the widespread activation shortly after appearance of the stimulus (from 100. ms) is consistent both with accounts of multi-pass processing, and distributed representations of categories. These methods provide an alternative to fMRI for fine-grained, large-scale investigations of the conceptual lexicon. © 2010 Elsevier Inc.

Android Malware Detection Using Parallel Machine Learning Classifiers

Mobile malware has continued to grow at an alarming rate despite on-going mitigation efforts. This has been much more prevalent on Android due to being an open platform that is rapidly overtaking other competing platforms in the mobile smart devices market. Recently, a new generation of Android malware families has emerged with advanced evasion capabilities which make them much more difficult to detect using conventional methods. This paper proposes and investigates a parallel machine learning based classification approach for early detection of Android malware. Using real malware samples and benign applications, a composite classification model is developed from parallel combination of heterogeneous classifiers. The empirical evaluation of the model under different combination schemes demonstrates its efficacy and potential to improve detection accuracy. More importantly, by utilizing several classifiers with diverse characteristics, their strengths can be harnessed not only for enhanced Android malware detection but also quicker white box analysis by means of the more interpretable constituent classifiers.

Choosing and using diversity indices: Insights for ecological applications from the German Biodiversity Exploratories

Biodiversity, a multidimensional property of natural systems, is difficult to quantify partly because of the multitude of indices proposed for this purpose. Indices aim to describe general properties of communities that allow us to compare different regions, taxa, and trophic levels. Therefore, they are of fundamental importance for environmental monitoring and conservation, although there is no consensus about which indices are more appropriate and informative. We tested several common diversity indices in a range of simple to complex statistical analyses in order to determine whether some were better suited for certain analyses than others. We used data collected around the focal plant Plantago lanceolata on 60 temperate grassland plots embedded in an agricultural landscape to explore relationships between the common diversity indices of species richness (S), Shannon's diversity (H'), Simpson's diversity (D1), Simpson's dominance (D2), Simpson's evenness (E), and Berger–Parker dominance (BP). We calculated each of these indices for herbaceous plants, arbuscular mycorrhizal fungi, aboveground arthropods, belowground insect larvae, and P. lanceolata molecular and chemical diversity. Including these trait-based measures of diversity allowed us to test whether or not they behaved similarly to the better studied species diversity. We used path analysis to determine whether compound indices detected more relationships between diversities of different organisms and traits than more basic indices. In the path models, more paths were significant when using H', even though all models except that with E were equally reliable. This demonstrates that while common diversity indices may appear interchangeable in simple analyses, when considering complex interactions, the choice of index can profoundly alter the interpretation of results. Data mining in order to identify the index producing the most significant results should be avoided, but simultaneously considering analyses using multiple indices can provide greater insight into the interactions in a system.