Self-organising coordination systems

Many research fields are pushing the engineering of large-scale, mobile, and open systems towards the adoption of techniques inspired by self-organisation: pervasive computing, but also distributed artificial intelligence, multi-agent systems, social networks, peer-topeer and grid architectures exploit adaptive techniques to make global system properties emerge in spite of the unpredictability of interactions and behaviour. Such a trend is visible also in coordination models and languages, whenever a coordination infrastructure needs to cope with managing interactions in highly dynamic and unpredictable environments. As a consequence, self-organisation can be regarded as a feasible metaphor to define a radically new conceptual coordination framework. The resulting framework defines a novel coordination paradigm, called self-organising coordination, based on the idea of spreading coordination media over the network, and charge them with services to manage interactions based on local criteria, resulting in the emergence of desired and fruitful global coordination properties of the system. Features like topology, locality, time-reactiveness, and stochastic behaviour play a key role in both the definition of such a conceptual framework and the consequent development of self-organising coordination services. According to this framework, the thesis presents several self-organising coordination techniques developed during the PhD course, mainly concerning data distribution in tuplespace-based coordination systems. Some of these techniques have been also implemented in ReSpecT, a coordination language for tuple spaces, based on logic tuples and reactions to events occurring in a tuple space. In addition, the key role played by simulation and formal verification has been investigated, leading to analysing how automatic verification techniques like probabilistic model checking can be exploited in order to formally prove the emergence of desired behaviours when dealing with coordination approaches based on self-organisation. To this end, a concrete case study is presented and discussed.

BioTuCSoN: biochemical extension of TuCSoN to support self-organising coordination

Starting from pervasive computing paradigm, we want to face the new system's requirements, concerning, mainly, self-organisation, situatedness and adaptivity, through the definition and execution of nature-inspired patterns. They are extracted by the study of dynamics in biological systems and we consider for their implementation the biochemical tuple spaces model. In particular, the aim of the thesis is to design and realize a first biochemical extension of TuCSoN (technology based on tuple spaces model) and, then, to verify its capabilities by means of a proper case study, that deals with local self-organisation and competition of services in an open and highly-dynamic environment.

UPLC-ESI-TOFMS-based metabolomics and gene expression dynamics inspector self-organizing metabolomic maps as tools for understanding the cellular response to ionizing radiation

Global transcriptomic and proteomic profiling platforms have yielded important insights into the complex response to ionizing radiation (IR). Nonetheless, little is known about the ways in which small cellular metabolite concentrations change in response to IR. Here, a metabolomics approach using ultraperformance liquid chromatography coupled with electrospray time-of-flight mass spectrometry was used to profile, over time, the hydrophilic metabolome of TK6 cells exposed to IR doses ranging from 0.5 to 8.0 Gy. Multivariate data analysis of the positive ions revealed dose- and time-dependent clustering of the irradiated cells and identified certain constituents of the water-soluble metabolome as being significantly depleted as early as 1 h after IR. Tandem mass spectrometry was used to confirm metabolite identity. Many of the depleted metabolites are associated with oxidative stress and DNA repair pathways. Included are reduced glutathione, adenosine monophosphate, nicotinamide adenine dinucleotide, and spermine. Similar measurements were performed with a transformed fibroblast cell line, BJ, and it was found that a subset of the identified TK6 metabolites were effective in IR dose discrimination. The GEDI (Gene Expression Dynamics Inspector) algorithm, which is based on self-organizing maps, was used to visualize dynamic global changes in the TK6 metabolome that resulted from IR. It revealed dose-dependent clustering of ions sharing the same trends in concentration change across radiation doses. "Radiation metabolomics," the application of metabolomic analysis to the field of radiobiology, promises to increase our understanding of cellular responses to stressors such as radiation.

Formation of temporal-feature maps by axonal propagation of synaptic learning

Computational maps are of central importance to a neuronal representation of the outside world. In a map, neighboring neurons respond to similar sensory features. A well studied example is the computational map of interaural time differences (ITDs), which is essential to sound localization in a variety of species and allows resolution of ITDs of the order of 10 μs. Nevertheless, it is unclear how such an orderly representation of temporal features arises. We address this problem by modeling the ontogenetic development of an ITD map in the laminar nucleus of the barn owl. We show how the owl's ITD map can emerge from a combined action of homosynaptic spike-based Hebbian learning and its propagation along the presynaptic axon. In spike-based Hebbian learning, synaptic strengths are modified according to the timing of pre- and postsynaptic action potentials. In unspecific axonal learning, a synapse's modification gives rise to a factor that propagates along the presynaptic axon and affects the properties of synapses at neighboring neurons. Our results indicate that both Hebbian learning and its presynaptic propagation are necessary for map formation in the laminar nucleus, but the latter can be orders of magnitude weaker than the former. We argue that the algorithm is important for the formation of computational maps, when, in particular, time plays a key role.

Exchange rate forecasting using a combined parametric and nonparametric self-organising modelling approach

Financial prediction has attracted a lot of interest due to the financial implications that the accurate prediction of financial markets can have. A variety of data driven modellingapproaches have been applied but their performance has produced mixed results. In this study we apply both parametric (neural networks with active neurons) and nonparametric (analog complexing) self-organisingmodelling methods for the daily prediction of the exchangerate market. We also propose acombinedapproach where the parametric and nonparametricself-organising methods are combined sequentially, exploiting the advantages of the individual methods with the aim of improving their performance. The combined method is found to produce promising results and to outperform the individual methods when tested with two exchangerates: the American Dollar and the Deutche Mark against the British Pound.

Beyond the topological map : developing alternate mappings in self organisation

The self organising map is a well established unsupervised
learning technique which is able to form sophisticated representations of an input data set. However, conventional Self Organising Map (SOM) algorithms are limited to the production of topological maps — that is, maps where distance between points on the map have a direct relationship to the Euclidean distance between the training vectors corresponding to those points.

It would be desirable to be able to create maps which form clusters on primitive attributes other than Euclidean distance; for example, clusters based upon orientation or shape. Such maps could provide a novel approach to pattern recognition tasks by providing a new method to associate groups of data.

In this paper, it is shown that the type of map produced by SOM algorithms is a direct consequence of the lateral connection strategy employed. Given this knowledge, a technique is required to establish the feasability of using an alternative lateral connection strategy. Such a technique is presented. Using this technique, it is possible to rule out lateral connection strategies that will not produce output states useful to the organisation process. This technique is demonstrated using conventional Laplacian interconnection as well as a number of novel interconnection strategies.

Analysing stillbirth data using dynamic self organizing maps

Even with the presence of modern obstetric care, stillbirth rate seems to stay stagnant or has even risen slightly in countries such as England and has become a significant public health concern [1]. In the light of current medical research, maternal risk factors such as diabetes and hypertensive disease were identified as possible risk factors and are taken into consideration in antenatal care. However, medical practitioners and researchers suspect possible relationships between trends in maternal demographics, antenatal care and pregnancy information of current stillbirth in consideration [2]. Although medical data and knowledge is available appropriate computing techniques to analyze the data may lead to identification of high risk groups. In this paper we use an unsupervised clustering technique called Growing Self organizing Map (GSOM) to analyse the stillbirth data and present patterns which can be important to medical researchers.

Growing self-organizing map for online continuous clustering

The internet age has fuelled an enormous explosion in the amount of information generated by humanity. Much of this information is transient in nature, created to be immediately consumed and built upon (or discarded). The field of data mining is surprisingly scant with algorithms that are geared towards the unsupervised knowledge extraction of such dynamic data streams. This chapter describes a new neural network algorithm inspired by self-organising maps. The new algorithm is a hybrid algorithm from the growing self-organising map (GSOM) and the cellular probabilistic self-organising map (CPSOM). The result is an algorithm which generates a dynamically growing feature map for the purpose of clustering dynamic data streams and tracking clusters as they evolve in the data stream.

Self-Organizing Maps of Molecular Descriptors for Sesquiterpene Lactones and Their Application to the Chemotaxonomy of the Asteraceae Family

The Asteraceae, one of the largest families among angiosperms, is chemically characterised by the production of sesquiterpene lactones (SLs). A total of 1,111 SLs, which were extracted from 658 species, 161 genera, 63 subtribes and 15 tribes of Asteraceae, were represented and registered in two dimensions in the SISTEMATX, an in-house software system, and were associated with their botanical sources. The respective 11 block of descriptors: Constitutional, Functional groups, BCUT, Atom-centred, 2D autocorrelations, Topological, Geometrical, RDF, 3D-MoRSE, GETAWAY and WHIM were used as input data to separate the botanical occurrences through self-organising maps. Maps that were generated with each descriptor divided the Asteraceae tribes, with total index values between 66.7% and 83.6%. The analysis of the results shows evident similarities among the Heliantheae, Helenieae and Eupatorieae tribes as well as between the Anthemideae and Inuleae tribes. Those observations are in agreement with systematic classifications that were proposed by Bremer, which use mainly morphological and molecular data, therefore chemical markers partially corroborate with these classifications. The results demonstrate that the atom-centred and RDF descriptors can be used as a tool for taxonomic classification in low hierarchical levels, such as tribes. Descriptors obtained through fragments or by the two-dimensional representation of the SL structures were sufficient to obtain significant results, and better results were not achieved by using descriptors derived from three-dimensional representations of SLs. Such models based on physico-chemical properties can project new design SLs, similar structures from literature or even unreported structures in two-dimensional chemical space. Therefore, the generated SOMs can predict the most probable tribe where a biologically active molecule can be found according Bremer classification.

A combined measure for quantifying and qualifying the topology preservation of growing self-organizing maps

The Self-OrganizingMap (SOM) is a neural network model that performs an ordered projection of a high dimensional input space in a low-dimensional topological structure. The process in which such mapping is formed is defined by the SOM algorithm, which is a competitive, unsupervised and nonparametric method, since it does not make any assumption about the input data distribution. The feature maps provided by this algorithm have been successfully applied for vector quantization, clustering and high dimensional data visualization processes. However, the initialization of the network topology and the selection of the SOM training parameters are two difficult tasks caused by the unknown distribution of the input signals. A misconfiguration of these parameters can generate a feature map of low-quality, so it is necessary to have some measure of the degree of adaptation of the SOM network to the input data model. The topologypreservation is the most common concept used to implement this measure. Several qualitative and quantitative methods have been proposed for measuring the degree of SOM topologypreservation, particularly using Kohonen's model. In this work, two methods for measuring the topologypreservation of the Growing Cell Structures (GCSs) model are proposed: the topographic function and the topology preserving map

Identifying the damage mechanisms in paper using the acoustic emission monitoring technique

This paper investigates the use of the acoustic emission (AE) monitoring technique for use in identifying the damage mechanisms present in paper associated with its production process. The microscopic structure of paper consists of a random mesh of paper fibres connected by hydrogen bonds. This implies the existence of two damage mechanisms, the failure of a fibre-fibre bond and the failure of a fibre. This paper describes a hybrid mathematical model which couples the mechanics of the mass-spring model to the acoustic wave propagation model for use in generating the acoustic signal emitted by complex structures of paper fibres under strain. The derivation of the mass-spring model can be found in [1,2], with details of the acoustic wave equation found in [3,4]. The numerical implementation of the vibro-acoustic model is discussed in detail with particular emphasis on the damping present in the numerical model. The hybrid model uses an implicit solver which intrinsically introduces artificial damping to the solution. The artificial damping is shown to affect the frequency response of the mass-spring model, therefore certain restrictions on the simulation time step must be enforced so that the model produces physically accurate results. The hybrid mathematical model is used to simulate small fibre networks to provide information on the acoustic response of each damage mechanism. The simulated AEs are then analysed using a continuous wavelet transform (CWT), described in [5], which provides a two dimensional time-frequency representation of the signal. The AEs from the two damage mechanisms show different characteristics in the CWT so that it is possible to define a fibre-fibre bond failure by the criteria listed below. The dominant frequency components of the AE must be at approximately 250 kHz or 750 kHz. The strongest frequency component may be at either approximately 250 kHz or 750 kHz. The duration of the frequency component at approximately 250 kHz is longer than that of the frequency component at approximately 750 kHz. Similarly, the criteria for identifying a fibre failure are given below. The dominant frequency component of the AE must be greater than 800 kHz. The duration of the dominant frequency component must be less than 5.00E-06 seconds. The dominant frequency component must be present at the front of the AE. Essentially, the failure of a fibre-fibre bond produces a low frequency wave and the failure of a fibre produces a high frequency pulse. Using this theoretical criteria, it is now possible to train an intelligent classifier such as the Self-Organising Map (SOM) [6] using the experimental data. First certain features must be extracted from the CWTs of the AEs for use in training the SOM. For this work, each CWT is divided into 200 windows of 5E-06s in duration covering a 100 kHz frequency range. The power ratio for each windows is then calculated and used as a feature. Having extracted the features from the AEs, the SOM can now be trained, but care is required so that the both damage mechanisms are adequately represented in the training set. This is an issue with paper as the failure of the fibre-fibre bonds is the prevalent damage mechanism. Once a suitable training set is found, the SOM can be trained and its performance analysed. For the SOM described in this work, there is a good chance that it will correctly classify the experimental AEs.

Multimodal semantic-associative collateral labelling and indexing of still images

A novel framework for multimodal semantic-associative collateral image labelling, aiming at associating image regions with textual keywords, is described. Both the primary image and collateral textual modalities are exploited in a cooperative and complementary fashion. The collateral content and context based knowledge is used to bias the mapping from the low-level region-based visual primitives to the high-level visual concepts defined in a visual vocabulary. We introduce the notion of collateral context, which is represented as a co-occurrence matrix, of the visual keywords, A collaborative mapping scheme is devised using statistical methods like Gaussian distribution or Euclidean distance together with collateral content and context-driven inference mechanism. Finally, we use Self Organising Maps to examine the classification and retrieval effectiveness of the proposed high-level image feature vector model which is constructed based on the image labelling results.

An empirical study of neighbourhood decay in Kohonen's self organizing map

In this paper, empirical results are presented which suggest that size and rate of decay of region size plays a much more significant role in the learning, and especially the development, of topographic feature maps. Using these results as a basis, a scheme for decaying region size during SOM training is proposed. The proposed technique provides near optimal training time. This scheme avoids the need for sophisticated learning gain decay schemes, and precludes the need for a priori knowledge of likely training times. This scheme also has some potential uses for continuous learning.