Learning the dynamics of the external world: Brain inspired learning for robotic applications

Humans have exceptional abilities to learn new skills, manipulate tools and objects, and interact with our environment. In order to be successful at these tasks, our brain has become exceptionally well adapted to learning to deal not only with the complex dynamics of our own limbs but also with novel dynamics in the external world. While learning of these dynamics includes learning the complex time-varying forces at the end of limbs through the updating of internal models, it must also include learning the appropriate mechanical impedance in order to stabilize both the limb and any objects contacted in the environment. This article reviews the field of human learning by examining recent experimental evidence about adaptation to novel unstable dynamics and explores how this knowledge about the brain and neuro-muscular system can expand the learning capabilities of robotics and prosthetics. © 2006.

Learning large-scale spatial relationships in a maze and effects of MK-801 on retrieval in the rhesus monkey

Monkeys have strong abilities to remember the visual properties of potential food sources for survival in the nature. The present study demonstrated the first observations of rhesus monkeys learning to solve complex spatial mazes in which routes were guid

Impedance control is selectively tuned to multiple directions of movement

Humans are able to learn tool-handling tasks, such as carving, demonstrating their competency to make movements in unstable environments with varied directions. When faced with a single direction of instability, humans learn to selectively co-contract their arm muscles tuning the mechanical stiffness of the limb end point to stabilize movements. This study examines, for the first time, subjects simultaneously adapting to two distinct directions of instability, a situation that may typically occur when using tools. Subjects learned to perform reaching movements in two directions, each of which had lateral instability requiring control of impedance. The subjects were able to adapt to these unstable interactions and switch between movements in the two directions; they did so by learning to selectively control the end-point stiffness counteracting the environmental instability without superfluous stiffness in other directions. This finding demonstrates that the central nervous system can simultaneously tune the mechanical impedance of the limbs to multiple movements by learning movement-specific solutions. Furthermore, it suggests that the impedance controller learns as a function of the state of the arm rather than a general strategy. © 2011 the American Physiological Society.

Bayesian Active Learning for Classification and Preference Learning

Information theoretic active learning has been widely studied for probabilistic models. For simple regression an optimal myopic policy is easily tractable. However, for other tasks and with more complex models, such as classification with nonparametric models, the optimal solution is harder to compute. Current approaches make approximations to achieve tractability. We propose an approach that expresses information gain in terms of predictive entropies, and apply this method to the Gaussian Process Classifier (GPC). Our approach makes minimal approximations to the full information theoretic objective. Our experimental performance compares favourably to many popular active learning algorithms, and has equal or lower computational complexity. We compare well to decision theoretic approaches also, which are privy to more information and require much more computational time. Secondly, by developing further a reformulation of binary preference learning to a classification problem, we extend our algorithm to Gaussian Process preference learning.

Model based learning of sigma points in unscented Kalman filtering

The unscented Kalman filter (UKF) is a widely used method in control and time series applications. The UKF suffers from arbitrary parameters necessary for sigma point placement, potentially causing it to perform poorly in nonlinear problems. We show how to treat sigma point placement in a UKF as a learning problem in a model based view. We demonstrate that learning to place the sigma points correctly from data can make sigma point collapse much less likely. Learning can result in a significant increase in predictive performance over default settings of the parameters in the UKF and other filters designed to avoid the problems of the UKF, such as the GP-ADF. At the same time, we maintain a lower computational complexity than the other methods. We call our method UKF-L. © 2011 Elsevier B.V.

Machine learning of functional class from phenotype data

Clare, A. and King R.D. (2002) Machine learning of functional class from phenotype data. Bioinformatics 18(1) 160-166

How Does the Cerebral Cortex Work? Developement, Learning, Attention, and 3D Vision by Laminar Circuits of Visual Cortex

A key goal of behavioral and cognitive neuroscience is to link brain mechanisms to behavioral functions. The present article describes recent progress towards explaining how the visual cortex sees. Visual cortex, like many parts of perceptual and cognitive neocortex, is organized into six main layers of cells, as well as characteristic sub-lamina. Here it is proposed how these layered circuits help to realize the processes of developement, learning, perceptual grouping, attention, and 3D vision through a combination of bottom-up, horizontal, and top-down interactions. A key theme is that the mechanisms which enable developement and learning to occur in a stable way imply properties of adult behavior. These results thus begin to unify three fields: infant cortical developement, adult cortical neurophysiology and anatomy, and adult visual perception. The identified cortical mechanisms promise to generalize to explain how other perceptual and cognitive processes work.

A Quantitative Evaluation of the AVITEWRITE Model of Handwriting Learning

Much sensory-motor behavior develops through imitation, as during the learning of handwriting by children. Such complex sequential acts are broken down into distinct motor control synergies, or muscle groups, whose activities overlap in time to generate continuous, curved movements that obey an intense relation between curvature and speed. The Adaptive Vector Integration to Endpoint (AVITEWRITE) model of Grossberg and Paine (2000) proposed how such complex movements may be learned through attentive imitation. The model suggest how frontal, parietal, and motor cortical mechanisms, such as difference vector encoding, under volitional control from the basal ganglia, interact with adaptively-timed, predictive cerebellar learning during movement imitation and predictive performance. Key psycophysical and neural data about learning to make curved movements were simulated, including a decrease in writing time as learning progresses; generation of unimodal, bell-shaped velocity profiles for each movement synergy; size scaling with isochrony, and speed scaling with preservation of the letter shape and the shapes of the velocity profiles; an inverse relation between curvature and tangential velocity; and a Two-Thirds Power Law relation between angular velocity and curvature. However, the model learned from letter trajectories of only one subject, and only qualitative kinematic comparisons were made with previously published human data. The present work describes a quantitative test of AVITEWRITE through direct comparison of a corpus of human handwriting data with the model's performance when it learns by tracing human trajectories. The results show that model performance was variable across subjects, with an average correlation between the model and human data of 89+/-10%. The present data from simulations using the AVITEWRITE model highlight some of its strengths while focusing attention on areas, such as novel shape learning in children, where all models of handwriting and learning of other complex sensory-motor skills would benefit from further research.

Large-Scale Neural Systems for Vision and Cognition

— Consideration of how people respond to the question What is this? has suggested new problem frontiers for pattern recognition and information fusion, as well as neural systems that embody the cognitive transformation of declarative information into relational knowledge. In contrast to traditional classification methods, which aim to find the single correct label for each exemplar (This is a car), the new approach discovers rules that embody coherent relationships among labels which would otherwise appear contradictory to a learning system (This is a car, that is a vehicle, over there is a sedan). This talk will describe how an individual who experiences exemplars in real time, with each exemplar trained on at most one category label, can autonomously discover a hierarchy of cognitive rules, thereby converting local information into global knowledge. Computational examples are based on the observation that sensors working at different times, locations, and spatial scales, and experts with different goals, languages, and situations, may produce apparently inconsistent image labels, which are reconciled by implicit underlying relationships that the network’s learning process discovers. The ARTMAP information fusion system can, moreover, integrate multiple separate knowledge hierarchies, by fusing independent domains into a unified structure. In the process, the system discovers cross-domain rules, inferring multilevel relationships among groups of output classes, without any supervised labeling of these relationships. In order to self-organize its expert system, the ARTMAP information fusion network features distributed code representations which exploit the model’s intrinsic capacity for one-to-many learning (This is a car and a vehicle and a sedan) as well as many-to-one learning (Each of those vehicles is a car). Fusion system software, testbed datasets, and articles are available from http://cns.bu.edu/techlab.

Biased ART: A Neural Architecture that Shifts Attention Toward Previously Disregarded Features Following an Incorrect Prediction

Memories in Adaptive Resonance Theory (ART) networks are based on matched patterns that focus attention on those portions of bottom-up inputs that match active top-down expectations. While this learning strategy has proved successful for both brain models and applications, computational examples show that attention to early critical features may later distort memory representations during online fast learning. For supervised learning, biased ARTMAP (bARTMAP) solves the problem of over-emphasis on early critical features by directing attention away from previously attended features after the system makes a predictive error. Small-scale, hand-computed analog and binary examples illustrate key model dynamics. Twodimensional simulation examples demonstrate the evolution of bARTMAP memories as they are learned online. Benchmark simulations show that featural biasing also improves performance on large-scale examples. One example, which predicts movie genres and is based, in part, on the Netflix Prize database, was developed for this project. Both first principles and consistent performance improvements on all simulation studies suggest that featural biasing should be incorporated by default in all ARTMAP systems. Benchmark datasets and bARTMAP code are available from the CNS Technology Lab Website: http://techlab.bu.edu/bART/.

Self-Supervised ARTMAP

Computational models of learning typically train on labeled input patterns (supervised learning), unlabeled input patterns (unsupervised learning), or a combination of the two (semisupervised learning). In each case input patterns have a fixed number of features throughout training and testing. Human and machine learning contexts present additional opportunities for expanding incomplete knowledge from formal training, via self-directed learning that incorporates features not previously experienced. This article defines a new self-supervised learning paradigm to address these richer learning contexts, introducing a neural network called self-supervised ARTMAP. Self-supervised learning integrates knowledge from a teacher (labeled patterns with some features), knowledge from the environment (unlabeled patterns with more features), and knowledge from internal model activation (self-labeled patterns). Self-supervised ARTMAP learns about novel features from unlabeled patterns without destroying partial knowledge previously acquired from labeled patterns. A category selection function bases system predictions on known features, and distributed network activation scales unlabeled learning to prediction confidence. Slow distributed learning on unlabeled patterns focuses on novel features and confident predictions, defining classification boundaries that were ambiguous in the labeled patterns. Self-supervised ARTMAP improves test accuracy on illustrative lowdimensional problems and on high-dimensional benchmarks. Model code and benchmark data are available from: http://techlab.bu.edu/SSART/.

Space, Time and Learning in the Hippocampus: How Fine Spatial and Temporal Scales Are Expanded into Population Codes for Behavioral Control

The hippocampus participates in multiple functions, including spatial navigation, adaptive timing, and declarative (notably, episodic) memory. How does it carry out these particular functions? The present article proposes that hippocampal spatial and temporal processing are carried out by parallel circuits within entorhinal cortex, dentate gyrus, and CA3 that are variations of the same circuit design. In particular, interactions between these brain regions transform fine spatial and temporal scales into population codes that are capable of representing the much larger spatial and temporal scales that are needed to control adaptive behaviors. Previous models of adaptively timed learning propose how a spectrum of cells tuned to brief but different delays are combined and modulated by learning to create a population code for controlling goal-oriented behaviors that span hundreds of milliseconds or even seconds. Here it is proposed how projections from entorhinal grid cells can undergo a similar learning process to create hippocampal place cells that can cover a space of many meters that are needed to control navigational behaviors. The suggested homology between spatial and temporal processing may clarify how spatial and temporal information may be integrated into an episodic memory.

Fuzzy ARTMAP: A Neural Network Architecture for Incremental Supervised Learning of Analog Multidimensional Maps

A new neural network architecture is introduced for incremental supervised learning of recognition categories and multidimensional maps in response to arbitrary sequences of analog or binary input vectors. The architecture, called Fuzzy ARTMAP, achieves a synthesis of fuzzy logic and Adaptive Resonance Theory (ART) neural networks by exploiting a close formal similarity between the computations of fuzzy subsethood and ART category choice, resonance, and learning. Fuzzy ARTMAP also realizes a new Minimax Learning Rule that conjointly minimizes predictive error and maximizes code compression, or generalization. This is achieved by a match tracking process that increases the ART vigilance parameter by the minimum amount needed to correct a predictive error. As a result, the system automatically learns a minimal number of recognition categories, or "hidden units", to met accuracy criteria. Category proliferation is prevented by normalizing input vectors at a preprocessing stage. A normalization procedure called complement coding leads to a symmetric theory in which the MIN operator (Λ) and the MAX operator (v) of fuzzy logic play complementary roles. Complement coding uses on-cells and off-cells to represent the input pattern, and preserves individual feature amplitudes while normalizing the total on-cell/off-cell vector. Learning is stable because all adaptive weights can only decrease in time. Decreasing weights correspond to increasing sizes of category "boxes". Smaller vigilance values lead to larger category boxes. Improved prediction is achieved by training the system several times using different orderings of the input set. This voting strategy can also be used to assign probability estimates to competing predictions given small, noisy, or incomplete training sets. Four classes of simulations illustrate Fuzzy ARTMAP performance as compared to benchmark back propagation and genetic algorithm systems. These simulations include (i) finding points inside vs. outside a circle; (ii) learning to tell two spirals apart; (iii) incremental approximation of a piecewise continuous function; and (iv) a letter recognition database. The Fuzzy ARTMAP system is also compared to Salzberg's NGE system and to Simpson's FMMC system.

Team-teaching for inclusive learning: Purposes, practices and perceptions of a team-teaching initiative in Irish post-primary schools

Background. Schools unequivocally privilege solo-teaching. This research seeks to enhance our understanding of team-teaching by examining how two teachers, working in the same classroom at the same time, might or might not contribute to the promotion of inclusive learning. There are well-established policy statements that encourage change and moves towards the use of team-teaching to promote greater inclusion of students with special educational needs in mainstream schools and mainstream classrooms. What is not so well established is the practice of team-teaching in post-primary settings, with little research conducted to date on how it can be initiated and sustained, and a dearth of knowledge on how it impacts upon the students and teachers involved. Research questions and aims. In light of the paucity and inconclusive nature of the research on team-teaching to date (Hattie, 2009), the orientating question in this study asks ‘To what extent, can the introduction of a formal team-teaching initiative enhance the quality of inclusive student learning and teachers’ learning at post-primary level?’ The framing of this question emerges from ongoing political, legal and educational efforts to promote inclusive education. The study has three main aims. The first aim of this study is to gather and represent the voices and experiences of those most closely involved in the introduction of team-teaching; students, teachers, principals and administrators. The second aim is to generate a theory-informed understanding of such collaborative practices and how they may best be implemented in the future. The third aim is to advance our understandings regarding the day-to-day, and moment-to-moment interactions, between teachers and students which enable or inhibit inclusive learning. Sample. In total, 20 team-teaching dyads were formed across seven project schools. The study participants were from two of the seven project schools, Ash and Oak. It involved eight teachers and 53 students, whose age ranged from 12-16 years old, with 4 teachers forming two dyads per school. In Oak there was a class of first years (n=11) with one dyad and a class of transition year students (n=24) with the other dyad. In Ash one class group (n=18) had two dyads. The subjects in which the dyads engaged were English and Mathematics. Method. This research adopted an interpretive paradigm. The duration of the fieldwork was from April 2007 to June 2008. Research methodologies included semi-structured interviews (n=44), classroom observation (n=20), attendance at monthly teacher meetings (n=6), questionnaires and other data gathering practices which included school documentation, assessment findings and joint examination of student work samples (n=4). Results. Team-teaching involves changing normative practices, and involves placing both demands and opportunities before those who occupy classrooms (teachers and students) and before those who determine who should occupy these classrooms (principals and district administrators). This research shows how team-teaching has the potential to promote inclusive learning, and when implemented appropriately, can impact positively upon the learning experiences of both teachers and students. The results are outlined in two chapters. In chapter four, Social Capital Theory is used in framing the data, the change process of bonding, bridging and linking, and in capturing what the collaborative action of team-teaching means, asks and offers teachers; within classes, between classes, between schools and within the wider educational community. In chapter five, Positioning Theory deductively assists in revealing the moment-to-moment, dynamic and inclusive learning opportunities, that are made available to students through team-teaching. In this chapter a number of vignettes are chosen to illustrate such learning opportunities. These two theories help to reveal the counter-narrative that team-teaching offers, regarding how both teachers and students teach and learn. This counter-narrative can extend beyond the field of special education and include alternatives to the manner in which professional development is understood, implemented, and sustained in schools and classrooms. Team-teaching repositions teachers and students to engage with one another in an atmosphere that capitalises upon and builds relational trust and shared cognition. However, as this research study has found, it is wise that the purposes, processes and perceptions of team-teaching are clear to all so that team-teaching can be undertaken by those who are increasingly consciously competent and not merely accidentally adequate. Conclusions. The findings are discussed in the context of the promotion of effective inclusive practices in mainstream settings. I believe that such promotion requires more nuanced understandings of what is being asked of, and offered to, teachers and students. Team-teaching has, and I argue will increasingly have, its place in the repertoire of responses that support effective inclusive learning. To capture and extend such practice requires theoretical frameworks that facilitate iterative journeys between research, policy and practice. Research to date on team-teaching has been too focused on outcomes over short timeframes and not focused enough on the process that is team-teaching. As a consequence team-teaching has been under-used, under-valued, under-theorised and generally not very well understood. Moving from classroom to staff room and district board room, theoretical frameworks used in this research help to travel with, and understand, the initiation, engagement and early consequences of team-teaching within and across the educational landscape. Therefore, conclusions from this study have implications for the triad of research, practice and policy development where efforts to change normative practices can be matched by understandings associated with what it means to try something new/anew, and what it means to say it made a positive difference.

The impact of learning context on the acquisition of sociopragmatic variation patterns in non-native speaker teachers of English

The study is a cross-linguistic, cross-sectional investigation of the impact of learning contexts on the acquisition of sociopragmatic variation patterns and the subsequent enactment of compound identities. The informants are 20 non-native speaker teachers of English from a range of 10 European countries. They are all primarily mono-contextual foreign language learners/users of English: however, they differ with respect to the length of time accumulated in a target language environment. This allows for three groups to be established – those who have accumulated 60 days or less; those with between 90 days and one year and the final group, all of whom have accumulated in excess of one year. In order to foster the dismantling of the monolith of learning context, both learning contexts under consideration – i.e. the foreign language context and submersion context are broken down into micro-contexts which I refer to as loci of learning. For the purpose of this study, two loci are considered: the institutional and the conversational locus. In order to make a correlation between the impact of learning contexts and loci of learning on the acquisition of sociopragmatic variation patterns, a two-fold study is conducted. The first stage is the completion of a highly detailed language contact profile (LCP) questionnaire. This provides extensive biographical information regarding language learning history and is a powerful tool in illuminating the intensity of contact with the L2 that learners experience in both contexts as well as shedding light on the loci of learning to which learners are exposed in both contexts. Following the completion of the LCP, the informants take part in two role plays which require the enactment of differential identities when engaged in a speech event of asking for advice. The enactment of identities then undergoes a strategic and linguistic analysis in order to investigate if and how differences in the enactment of compound identities are indexed in language. Results indicate that learning context has a considerable impact not only on how identity is indexed in language, but also on the nature of identities enacted. Informants with very low levels of crosscontextuality index identity through strategic means – i.e. levels of directness and conventionality; however greater degrees of cross-contextuality give rise to the indexing of differential identities linguistically by means of speaker/hearer orientation and (non-) solidary moves. When it comes to the nature of identity enacted, it seems that more time spent in intense contact with native speakers in a range of loci of learning allows learners to enact their core identity; whereas low levels of contact with over-exposure to the institutional locus of learning fosters the enactment of generic identities.