Clustering quality and topology preservation in fast learning SOMs

The Self-Organizing Map (SOM) is a popular unsupervised neural network able to provide effective clustering and data visualization for data represented in multidimensional input spaces. In this paper, we describe Fast Learning SOM (FLSOM) which adopts a learning algorithm that improves the performance of the standard SOM with respect to the convergence time in the training phase. We show that FLSOM also improves the quality of the map by providing better clustering quality and topology preservation of multidimensional input data. Several tests have been carried out on different multidimensional datasets, which demonstrate better performances of the algorithm in comparison with the original SOM.

A new RBF neural network with boundary value constraints

We present a novel topology of the radial basis function (RBF) neural network, referred to as the boundary value constraints (BVC)-RBF, which is able to automatically satisfy a set of BVC. Unlike most existing neural networks whereby the model is identified via learning from observational data only, the proposed BVC-RBF offers a generic framework by taking into account both the deterministic prior knowledge and the stochastic data in an intelligent manner. Like a conventional RBF, the proposed BVC-RBF has a linear-in-the-parameter structure, such that it is advantageous that many of the existing algorithms for linear-in-the-parameters models are directly applicable. The BVC satisfaction properties of the proposed BVC-RBF are discussed. Finally, numerical examples based on the combined D-optimality-based orthogonal least squares algorithm are utilized to illustrate the performance of the proposed BVC-RBF for completeness.

A forward constrained selection algorithm for probabilistic neural network

A new probabilistic neural network (PNN) learning algorithm based on forward constrained selection (PNN-FCS) is proposed. An incremental learning scheme is adopted such that at each step, new neurons, one for each class, are selected from the training samples arid the weights of the neurons are estimated so as to minimize the overall misclassification error rate. In this manner, only the most significant training samples are used as the neurons. It is shown by simulation that the resultant networks of PNN-FCS have good classification performance compared to other types of classifiers, but much smaller model sizes than conventional PNN.

Pattern separation in digital learning nets

Pattern separation is a new technique in digital learning networks which can be used to detect state conflicts. This letter describes pattern separation in a simple single-layer network, and an application of the technique in networks with feedback.

Adaptive modelling with tunable RBF network using multi-innovation RLS algorithm assisted by swarm intelligence

In this paper, we propose a new on-line learning algorithm for the non-linear system identification: the swarm intelligence aided multi-innovation recursive least squares (SI-MRLS) algorithm. The SI-MRLS algorithm applies the particle swarm optimization (PSO) to construct a flexible radial basis function (RBF) model so that both the model structure and output weights can be adapted. By replacing an insignificant RBF node with a new one based on the increment of error variance criterion at every iteration, the model remains at a limited size. The multi-innovation RLS algorithm is used to update the RBF output weights which are known to have better accuracy than the classic RLS. The proposed method can produces a parsimonious model with good performance. Simulation result are also shown to verify the SI-MRLS algorithm.

B-spline neural network based digital baseband predistorter solution using the inverse of De Boor algorithm

In this paper a new nonlinear digital baseband predistorter design is introduced based on direct learning, together with a new Wiener system modeling approach for the high power amplifiers (HPA) based on the B-spline neural network. The contribution is twofold. Firstly, by assuming that the nonlinearity in the HPA is mainly dependent on the input signal amplitude the complex valued nonlinear static function is represented by two real valued B-spline neural networks, one for the amplitude distortion and another for the phase shift. The Gauss-Newton algorithm is applied for the parameter estimation, in which the De Boor recursion is employed to calculate both the B-spline curve and the first order derivatives. Secondly, we derive the predistorter algorithm calculating the inverse of the complex valued nonlinear static function according to B-spline neural network based Wiener models. The inverse of the amplitude and phase shift distortion are then computed and compensated using the identified phase shift model. Numerical examples have been employed to demonstrate the efficacy of the proposed approaches.

Action observation and robotic agents: learning and anthropomorphism

The ‘action observation network’ (AON), which is thought to translate observed actions into motor codes required for their execution, is biologically tuned: it responds more to observation of human, than non-human, movement. This biological specificity has been taken to support the hypothesis that the AON underlies various social functions, such as theory of mind and action understanding, and that, when it is active during observation of non-human agents like humanoid robots, it is a sign of ascription of human mental states to these agents. This review will outline evidence for biological tuning in the AON, examining the features which generate it, and concluding that there is evidence for tuning to both the form and kinematic profile of observed movements, and little evidence for tuning to belief about stimulus identity. It will propose that a likely reason for biological tuning is that human actions, relative to non-biological movements, have been observed more frequently while executing corresponding actions. If the associative hypothesis of the AON is correct, and the network indeed supports social functioning, sensorimotor experience with non-human agents may help us to predict, and therefore interpret, their movements.

Multiscale evolving complex network model of functional connectivity in neuronal cultures

Cultures of cortical neurons grown on multielectrode arrays exhibit spontaneous, robust and recurrent patterns of highly synchronous activity called bursts. These bursts play a crucial role in the development and topological selforganization of neuronal networks. Thus, understanding the evolution of synchrony within these bursts could give insight into network growth and the functional processes involved in learning and memory. Functional connectivity networks can be constructed by observing patterns of synchrony that evolve during bursts. To capture this evolution, a modelling approach is adopted using a framework of emergent evolving complex networks and, through taking advantage of the multiple time scales of the system, aims to show the importance of sequential and ordered synchronization in network function.

Simulated annealing technique for fast learning of SOM networks

The Self-Organizing Map (SOM) is a popular unsupervised neural network able to provide effective clustering and data visualization for multidimensional input datasets. In this paper, we present an application of the simulated annealing procedure to the SOM learning algorithm with the aim to obtain a fast learning and better performances in terms of quantization error. The proposed learning algorithm is called Fast Learning Self-Organized Map, and it does not affect the easiness of the basic learning algorithm of the standard SOM. The proposed learning algorithm also improves the quality of resulting maps by providing better clustering quality and topology preservation of input multi-dimensional data. Several experiments are used to compare the proposed approach with the original algorithm and some of its modification and speed-up techniques.

Mid-term evaluation of African Technology Policy Studies Network Phase VI Strategic Plan: for the period January 1, 2009 to December 31, 2010

The African Technology Policy Studies Network (ATPS) is a multidisciplinary network of researchers, private sector actors, policymakers and civil society. ATPS has the vision to become the leading international centre of excellence and reference in science, technology and innovation (STI) systems research, training and capacity building, communication and sensitization, knowledge brokerage, policy advocacy and outreach in Africa. It has a Regional Secretariat in Nairobi Kenya, and operates through national chapters in 29 countries (including 27 in Africa and two Chapters in the United Kingdom and USA for Africans in the Diaspora) with an expansion plan to cover the entire continent by 2015. The ATPS Phase VI Strategic Plan aims to improve the understanding and functioning of STI processes and systems to strengthen the learning capacity, social responses, and governance of STI for addressing Africa's development challenges, with a specific focus on the Millennium Development Goals (MDGs). A team of external evaluators carried out a midterm review to assess the effectiveness and efficiency of the implementation of the Strategic Plan for the period January 1, 2009 to December 31, 2010. The evaluation methodology involved multiple quantitative and qualitative methods to assess the qualitative and quantitative inputs (human resources, financial resources, time, etc.) into ATPS activities (both thematic and facilitative) and their tangible and intangible outputs, outcomes and impacts. Methods included a questionnaire survey of ATPS members and stakeholders, key informant interviews, and focus group discussions (FGDs) with members in six countries. Effectiveness of Programmes Under all six strategic goals, very good progress has been made towards planned outputs and outcomes. This is evidenced by key performance indicators (KPIs) generated from desk review, ratings from the survey respondents, and the themes that run through the FGDs. Institutional and Programme Cost Effectiveness Institutional Effectiveness: assessment of institutional effectiveness suggests that adequate management frameworks are in place and are being used effectively and transparently. Also technical and financial accounting mechanisms are being followed in accordance with grant agreements and with global good practice. This is evidenced by KPIs generated from desk review. Programme Cost Effectiveness: assessment of cost-effectiveness of execution of programmes shows that organisational structure is efficient, delivering high quality, relevant research at relatively low cost by international standards. The evidence includes KPIs from desk review: administrative costs to programme cost ratio has fallen steadily, to around 10%; average size of research grants is modest, without compromising quality. There is high level of pro bono input by ATPS members. ATPS Programmes Strategic Evaluation ATPS research and STI related activities are indeed unique and well aligned with STI issues and needs facing Africa and globally. The multi-disciplinary and trans-boundary nature of the research activities are creating a unique group of research scientists. The ATPS approach to research and STI issues is paving the way for the so called Third Generation University (3GU). Understanding this unique positioning, an increasing number of international multilateral agencies are seeking partnership with ATPS. ATPS is seeing an increasing level of funding commitments by Donor Partners. Recommendations for ATPS Continued Growth and Effectiveness On-going reform of ATPS administrative structure to continue The on-going reforms that have taken place within the Board, Regional Secretariat, and at the National Chapter coordination levels are welcomed. Such reform should continue until fully functional corporate governance policy and practices are fully established and implemented across the ATPS governance structures. This will further strengthen ATPS to achieve the vision of being the leading STI policy brokerage organization in Africa. Although training in corporate governance has been carried out for all sectors of ATPS leadership structure in recent time, there is some evidence that these systems have not yet been fully implemented effectively within all the governance structures of the organization, especially at the Board and National chapter levels. Future training should emphasize practical application with exercises relevant to ATPS leadership structure from the Board to the National Chapter levels. Training on Transformational Leadership - Leading a Change Though a subject of intense debate amongst economists and social scientists, it is generally agreed that cultural mindsets and attitudes could enhance and/or hinder organizational progress. ATPS’s vision demands transformational leadership skills amongst its leaders from the Board members to the National Chapter Coordinators. To lead such a change, ATPS leaders must understand and avoid personal and cultural mindsets and value systems that hinder change, while embracing those that enhance it. It requires deliberate assessment of cultural, behavioural patterns that could hinder progress and the willingness to be recast into cultural and personal habits that make for progress. Improvement of relationship amongst the Board, Secretariat, and National Chapters A large number of ATPS members and stakeholders feel they do not have effective communications and/or access to Board, National Chapter Coordinators and Regional Secretariat activities. Effort should be made to improve the implementation of ATPS communication strategy to improve on information flows amongst the ATPS management and the members. The results of the survey and the FGDs suggest that progress has been made during the past two years in this direction, but more could be done to ensure effective flow of pertinent information to members following ATPS communications channels. Strategies for Increased Funding for National Chapters There is a big gap between the fundraising skills of the Regional Secretariat and those of the National Coordinators. In some cases, funds successfully raised by the Secretariat and disbursed to national chapters were not followed up with timely progress and financial reports by some national chapters. Adequate training in relevant skills required for effective interactions with STI key policy players should be conducted regularly for National Chapter coordinators and ATPS members. The ongoing training in grant writing should continue and be made continent-wide if funding permits. Funding of National Chapters should be strategic such that capacity in a specific area of research is built which, with time, will not only lead to a strong research capacity in that area, but also strengthen academic programmes. For example, a strong climate change programme is emerging at University of Nigeria Nsukka (UNN), with strong collaborations with Universities from neighbouring States. Strategies to Increase National Government buy-in and support for STI Translating STI research outcomes into policies requires a great deal of emotional intelligence, skills which are often lacking in the first and second generation universities. In the epoch of the science-based or 2GUs, governments were content with universities carrying out scientific research and providing scientific education. Now they desire to see universities as incubators of new science- or technology-based commercial activities, whether by existing firms or start-ups. Hence, governments demand that universities take an active and leading role in the exploitation of their knowledge and they are willing to make funds available to support such activities. Thus, for universities to gain the attention of national leadership they must become centres of excellence and explicit instruments of economic development in the knowledge-based economy. The universities must do this while working collaboratively with government departments, parastatals, and institutions and dedicated research establishments. ATPS should anticipate these shifting changes and devise programmes to assist both government and universities to relate effectively. New administrative structures in member organizations to sustain and manage the emerging STI multidisciplinary teams Second Generation universities (2GUs) tend to focus on pure science and often do not regard the application of their know-how as their task. In contrast, Third Generation Universities (3GUs) objectively stimulate techno-starters – students or academics – to pursue the exploitation or commercialisation of the knowledge they generate. They view this as being equal in importance to the objectives of scientific research and education. Administratively, research in the 2GU era was mainly monodisciplinary and departments were structured along disciplines. The emerging interdisciplinary scientific teams with focus on specific research areas functionally work against the current mono-disciplinary faculty-based, administrative structure of 2GUs. For interdisciplinary teams, the current faculty system is an obstacle. There is a need for new organisational forms for university management that can create responsibilities for the task of know-how exploitation. ATPS must anticipate this and begin to strategize solutions for their member institutions to transition to 3Gus administrative structure, otherwise ATPS growth will plateau, and progress achieved so far may be stunted.

Modular non-computational-connectionist-hybrid neural network approach to robotic systems

Spiking neural networks are usually limited in their applications due to their complex mathematical models and the lack of intuitive learning algorithms. In this paper, a simpler, novel neural network derived from a leaky integrate and fire neuron model, the ‘cavalcade’ neuron, is presented. A simulation for the neural network has been developed and two basic learning algorithms implemented within the environment. These algorithms successfully learn some basic temporal and instantaneous problems. Inspiration for neural network structures from these experiments are then taken and applied to process sensor information so as to successfully control a mobile robot.

Digital predistorter design using B-Spline neural network and inverse of De Boor algorithm

This contribution introduces a new digital predistorter to compensate serious distortions caused by memory high power amplifiers (HPAs) which exhibit output saturation characteristics. The proposed design is based on direct learning using a data-driven B-spline Wiener system modeling approach. The nonlinear HPA with memory is first identified based on the B-spline neural network model using the Gauss-Newton algorithm, which incorporates the efficient De Boor algorithm with both B-spline curve and first derivative recursions. The estimated Wiener HPA model is then used to design the Hammerstein predistorter. In particular, the inverse of the amplitude distortion of the HPA's static nonlinearity can be calculated effectively using the Newton-Raphson formula based on the inverse of De Boor algorithm. A major advantage of this approach is that both the Wiener HPA identification and the Hammerstein predistorter inverse can be achieved very efficiently and accurately. Simulation results obtained are presented to demonstrate the effectiveness of this novel digital predistorter design.

The plasticity of the mirror system: how reward learning modulates cortical motor simulation of others

Cortical motor simulation supports the understanding of others' actions and intentions. This mechanism is thought to rely on the mirror neuron system (MNS), a brain network that is active both during action execution and observation. Indirect evidence suggests that alpha/beta suppression, an electroencephalographic (EEG) index of MNS activity, is modulated by reward. In this study we aimed to test the plasticity of the MNS by directly investigating the link between alpha/beta suppression and reward. 40 individuals from a general population sample took part in an evaluative conditioning experiment, where different neutral faces were associated with high or low reward values. In the test phase, EEG was recorded while participants viewed videoclips of happy expressions made by the conditioned faces. Alpha/beta suppression (identified using event-related desynchronisation of specific independent components) in response to rewarding faces was found to be greater than for non-rewarding faces. This result provides a mechanistic insight into the plasticity of the MNS and, more generally, into the role of reward in modulating physiological responses linked to empathy.

Discovering biomarkers for antidepressant response: protocol from the Canadian biomarker integration network in depression (CAN-BIND) and clinical characteristics of the first patient cohort

Background Major Depressive Disorder (MDD) is among the most prevalent and disabling medical conditions worldwide. Identification of clinical and biological markers (“biomarkers”) of treatment response could personalize clinical decisions and lead to better outcomes. This paper describes the aims, design, and methods of a discovery study of biomarkers in antidepressant treatment response, conducted by the Canadian Biomarker Integration Network in Depression (CAN-BIND). The CAN-BIND research program investigates and identifies biomarkers that help to predict outcomes in patients with MDD treated with antidepressant medication. The primary objective of this initial study (known as CAN-BIND-1) is to identify individual and integrated neuroimaging, electrophysiological, molecular, and clinical predictors of response to sequential antidepressant monotherapy and adjunctive therapy in MDD. Methods CAN-BIND-1 is a multisite initiative involving 6 academic health centres working collaboratively with other universities and research centres. In the 16-week protocol, patients with MDD are treated with a first-line antidepressant (escitalopram 10–20 mg/d) that, if clinically warranted after eight weeks, is augmented with an evidence-based, add-on medication (aripiprazole 2–10 mg/d). Comprehensive datasets are obtained using clinical rating scales; behavioural, dimensional, and functioning/quality of life measures; neurocognitive testing; genomic, genetic, and proteomic profiling from blood samples; combined structural and functional magnetic resonance imaging; and electroencephalography. De-identified data from all sites are aggregated within a secure neuroinformatics platform for data integration, management, storage, and analyses. Statistical analyses will include multivariate and machine-learning techniques to identify predictors, moderators, and mediators of treatment response. Discussion From June 2013 to February 2015, a cohort of 134 participants (85 outpatients with MDD and 49 healthy participants) has been evaluated at baseline. The clinical characteristics of this cohort are similar to other studies of MDD. Recruitment at all sites is ongoing to a target sample of 290 participants. CAN-BIND will identify biomarkers of treatment response in MDD through extensive clinical, molecular, and imaging assessments, in order to improve treatment practice and clinical outcomes. It will also create an innovative, robust platform and database for future research.