Bregman divergences for growing hierarchical self-organizing networks

A Kernel Bayesian Adaptive Resonance Theory with A Topological Structure

This paper attempts to solve the typical problems of self-organizing growing network models, i.e. (a) an influence of the order of input data on the self-organizing ability, (b) an instability to high-dimensional data and an excessive sensitivity to noise, and (c) an expensive computational cost by integrating Kernel Bayes Rule (KBR) and Correntropy-Induced Metric (CIM) into Adaptive Resonance Theory (ART) framework. KBR performs a covariance-free Bayesian computation which is able to maintain a fast and stable computation. CIM is a generalized similarity measurement which can maintain a high-noise reduction ability even in a high-dimensional space. In addition, a Growing Neural Gas (GNG)-based topology construction process is integrated into the ART framework to enhance its self-organizing ability. The simulation experiments with synthetic and real-world datasets show that the proposed model has an outstanding stable self-organizing ability for various test environments.

Foreground Detection by Competitive Learning for Varying Input Distributions

One of the most important challenges in computer vision applications is the background modeling, especially when the background is dynamic and the input distribution might not be stationary, i.e. the distribution of the input data could change with time (e.g. changing illuminations, waving trees, water, etc.). In this work, an unsupervised learning neural network is proposed which is able to cope with progressive changes in the input distribution. It is based on a dual learning mechanism which manages the changes of the input distribution separately from the cluster detection. The proposal is adequate for scenes where the background varies slowly. The performance of the method is tested against several state-of-the-art foreground detectors both quantitatively and qualitatively, with favorable results.

Learning Topologies with the Growing Neural Forest

In this work, a novel self-organizing model called growing neural forest (GNF) is presented. It is based on the growing neural gas (GNG), which learns a general graph with no special provisions for datasets with separated clusters. On the contrary, the proposed GNF learns a set of trees so that each tree represents a connected cluster of data. High dimensional datasets often contain large empty regions among clusters, so this proposal is better suited to them than other self-organizing models because it represents these separated clusters as connected components made of neurons. Experimental results are reported which show the self-organization capabilities of the model. Moreover, its suitability for unsupervised clustering and foreground detection applications is demonstrated. In particular, the GNF is shown to correctly discover the connected component structure of some datasets. Moreover, it outperforms some well-known foreground detectors both in quantitative and qualitative terms.

Nature Inspired Computing: An Overview and Some Future Directions

This paper presents an overview of significant advances made in the emerging field of nature-inspired computing (NIC) with a focus on the physics- and biology-based approaches and algorithms. A parallel development in the past two decades has been the emergence of the field of computational intelligence (CI) consisting primarily of the three fields of neural networks, evolutionary computing and fuzzy logic. It is observed that NIC and CI intersect. The authors advocate and foresee more cross-fertilisation of the two emerging fields.

Discrimination of schizophrenia auditory hallucinators by machine learning of resting-state functional MRI

Auditory hallucinations (AH) are a symptom that is most often associated with schizophrenia, but patients with other neuropsychiatric conditions, and even a small percentage of healthy individuals, may also experience AH. Elucidating the neural mechanisms underlying AH in schizophrenia may offer insight into the pathophysiology associated with AH more broadly across multiple neuropsychiatric disease conditions. In this paper, we address the problem of classifying schizophrenia patients with and without a history of AH, and healthy control (HC) subjects. To this end, we performed feature extraction from resting state functional magnetic resonance imaging (rsfMRI) data and applied machine learning classifiers, testing two kinds of neuroimaging features: (a) functional connectivity (FC) measures computed by lattice auto-associative memories (LAAM), and (b) local activity (LA) measures, including regional homogeneity (ReHo) and fractional amplitude of low frequency fluctuations (fALFF). We show that it is possible to perform classification within each pair of subject groups with high accuracy. Discrimination between patients with and without lifetime AH was highest, while discrimination between schizophrenia patients and HC participants was worst, suggesting that classification according to the symptom dimension of AH may be more valid than discrimination on the basis of traditional diagnostic categories. FC measures seeded in right Heschl's gyrus (RHG) consistently showed stronger discriminative power than those seeded in left Heschl's gyrus (LHG), a finding that appears to support AH models focusing on right hemisphere abnormalities. The cortical brain localizations derived from the features with strong classification performance are consistent with proposed AH models, and include left inferior frontal gyrus (IFG), parahippocampal gyri, the cingulate cortex, as well as several temporal and prefrontal cortical brain regions. Overall, the observed findings suggest that computational intelligence approaches can provide robust tools for uncovering subtleties in complex neuroimaging data, and have the potential to advance the search for more neuroscience-based criteria for classifying mental illness in psychiatry research.

Principal polynomial analysis

This paper presents a new framework for manifold learning based on a sequence of principal polynomials that capture the possibly nonlinear nature of the data. The proposed Principal Polynomial Analysis (PPA) generalizes PCA by modeling the directions of maximal variance by means of curves, instead of straight lines. Contrarily to previous approaches, PPA reduces to performing simple univariate regressions, which makes it computationally feasible and robust. Moreover, PPA shows a number of interesting analytical properties. First, PPA is a volume-preserving map, which in turn guarantees the existence of the inverse. Second, such an inverse can be obtained in closed form. Invertibility is an important advantage over other learning methods, because it permits to understand the identified features in the input domain where the data has physical meaning. Moreover, it allows to evaluate the performance of dimensionality reduction in sensible (input-domain) units. Volume preservation also allows an easy computation of information theoretic quantities, such as the reduction in multi-information after the transform. Third, the analytical nature of PPA leads to a clear geometrical interpretation of the manifold: it allows the computation of Frenet-Serret frames (local features) and of generalized curvatures at any point of the space. And fourth, the analytical Jacobian allows the computation of the metric induced by the data, thus generalizing the Mahalanobis distance. These properties are demonstrated theoretically and illustrated experimentally. The performance of PPA is evaluated in dimensionality and redundancy reduction, in both synthetic and real datasets from the UCI repository.