Downloads & Free Reading Options - Results

NOTE: Unfortunately the recording of this talk was cut short by 5 minutes because of over runs earlier in the conference program. Tensor-Networks-based Learning of Probabilistic Cellular Automata Dynamics by Heitor Casagrande @QTMLConference

One of the key challenges in applying reinforcement learning to complex robotic control tasks is the need to gather large amounts of experience in order to find an effective policy for the task at hand. Model-based reinforcement learning can achieve good sample efficiency, but requires the ability to learn a model of the dynamics that is good enough to learn an effective policy. In this work, we develop a model-based reinforcement learning algorithm that combines prior knowledge from previous tasks with online adaptation of the dynamics model. These two ingredients enable highly sample-efficient learning even in regimes where estimating the true dynamics is very difficult, since the online model adaptation allows the method to locally compensate for unmodeled variation in the dynamics. We encode the prior experience into a neural network dynamics model, adapt it online by progressively refitting a local linear model of the dynamics, and use model predictive control to plan under these dynamics. Our experimental results show that this approach can be used to solve a variety of complex robotic manipulation tasks in just a single attempt, using prior data from other manipulation behaviors.

We apply a general theory describing the dynamics of supervised learning in layered neural networks in the regime where the size p of the training set is proportional to the number of inputs N, as developed in a previous paper, to several choices of learning rules. In the case of (on-line and batch) Hebbian learning, where a direct exact solution is possible, we show that our theory provides exact results at any time in many different verifiable cases. For non-Hebbian learning rules, such as Perceptron and AdaTron, we find very good agreement between the predictions of our theory and numerical simulations. Finally, we derive three approximation schemes aimed at eliminating the need to solve a functional saddle-point equation at each time step, and assess their performance. The simplest of these schemes leads to a fully explicit and relatively simple non-linear diffusion equation for the joint field distribution, which already describes the learning dynamics surprisingly well over a wide range of parameters.

We present a new model for the dynamics of the presynaptic intracellular calcium concentration in neurons evoked by various stimulation protocols. The aim of the model is twofold: We want to discuss the calcium transients during and after specific stimulation protocols as they are used to induce long-term-depression and long-term-potentiation. In addition we would like to provide a general tool which allows the comparison of different calcium experiments. This may help to draw conclusions on a wider base in future.

Reading has been perceived as a dynamic transactional process wherein readers negotiate meaning with writers by virtue of their prior knowledge. The act of meaning construction is realized primarily through the exploration of intertextual links that connect various sources of texts the readers have composed or experienced. Readers utilize intertextuality as an inquiry tool to build their own perspectives of the ideas and thoughts communicated in the texts. Recent inquiries corroborate the possibility of investigating intertextuality by analyzing reader response to the readings without purposefully establishing an "intertextuality rich environment." A study investigated the intertextual dynamics of the reading and inquiry processes of 23 advanced adult readers at the graduate levels of education in an authentic collaborative learning context that capitalizes on an online course listserv for open discussions. The research questions this study sought to answer are: What are the intertextual intentions of advanced adult readers? What are the intertextual loci that ground their intertextual processes? and What are the intertextual strategies readers employ to construct the meanings of the texts? This paper first discusses the concepts of text and intertextuality to formulate a theoretical frame to locate the investigation. The paper then provides a detailed analysis of readers' reflections and responses to address the three research questions. It then delineates a preliminary model of intertextual dynamics and examines the advantages of the listserv format in fostering intertextual inquiry. Includes a figure and 2 notes. Contains 41 references and cites 17 materials. (NKA)

This study used a longitudinal qualitative approach to study children's acquisition of scripts for school learning in New Zealand. Participating in the study were four boys and four girls, 5 years of age. Children were observed in regular class activities in their first term in a new entrant classroom and 1 year later. Informal and semi-structured interviews were conducted to elicit children's perceptions of learning tasks. The theoretical constructs of scripts, metacognition, and peer scaffolding provided the theoretical framework for the data reduction process. Results indicated that school beginners viewed learning in terms of procedural matters and classroom routines. While the children's focus on procedures seemed to limit their awareness of learning, it did appear to facilitate their attempt to give and receive scaffolding. There was little awareness of the self as learner, and learning as a concept was largely associated with the teacher. In their second year, children's perceptions of learning reflected changes in classroom dynamics and teaching styles and a greater awareness of learning content and strategies. The findings suggested that children look for patterns in school and construct scripts for school learning which reflect the changing dynamics of classrooms, and that peer interactions provide opportunities for developing relationships which support emergent forms of scaffolding, including metacognitive assistance to peers. Results suggested that teaching practices need to support the construction of scripts for learning and not simply scripts for routines and procedures. (Contains 25 references.) (KDFB)

This study used a longitudinal qualitative approach to study children's acquisition of scripts for school learning in New Zealand. Participating in the study were four boys and four girls, 5 years of age. Children were observed in regular class activities in their first term in a new entrant classroom and 1 year later. Informal and semi-structured interviews were conducted to elicit children's perceptions of learning tasks. The theoretical constructs of scripts, metacognition, and peer scaffolding provided the theoretical framework for the data reduction process. Results indicated that school beginners viewed learning in terms of procedural matters and classroom routines. While the children's focus on procedures seemed to limit their awareness of learning, it did appear to facilitate their attempt to give and receive scaffolding. There was little awareness of the self as learner, and learning as a concept was largely associated with the teacher. In their second year, children's perceptions of learning reflected changes in classroom dynamics and teaching styles and a greater awareness of learning content and strategies. The findings suggested that children look for patterns in school and construct scripts for school learning which reflect the changing dynamics of classrooms, and that peer interactions provide opportunities for developing relationships which support emergent forms of scaffolding, including metacognitive assistance to peers. Results suggested that teaching practices need to support the construction of scripts for learning and not simply scripts for routines and procedures. (Contains 25 references.) (KDFB)

This study used a longitudinal qualitative approach to study children's acquisition of scripts for school learning in New Zealand. Participating in the study were four boys and four girls, 5 years of age. Children were observed in regular class activities in their first term in a new entrant classroom and 1 year later. Informal and semi-structured interviews were conducted to elicit children's perceptions of learning tasks. The theoretical constructs of scripts, metacognition, and peer scaffolding provided the theoretical framework for the data reduction process. Results indicated that school beginners viewed learning in terms of procedural matters and classroom routines. While the children's focus on procedures seemed to limit their awareness of learning, it did appear to facilitate their attempt to give and receive scaffolding. There was little awareness of the self as learner, and learning as a concept was largely associated with the teacher. In their second year, children's perceptions of learning reflected changes in classroom dynamics and teaching styles and a greater awareness of learning content and strategies. The findings suggested that children look for patterns in school and construct scripts for school learning which reflect the changing dynamics of classrooms, and that peer interactions provide opportunities for developing relationships which support emergent forms of scaffolding, including metacognitive assistance to peers. Results suggested that teaching practices need to support the construction of scripts for learning and not simply scripts for routines and procedures. (Contains 25 references.) (KDFB)

This study used a longitudinal qualitative approach to study children's acquisition of scripts for school learning in New Zealand. Participating in the study were four boys and four girls, 5 years of age. Children were observed in regular class activities in their first term in a new entrant classroom and 1 year later. Informal and semi-structured interviews were conducted to elicit children's perceptions of learning tasks. The theoretical constructs of scripts, metacognition, and peer scaffolding provided the theoretical framework for the data reduction process. Results indicated that school beginners viewed learning in terms of procedural matters and classroom routines. While the children's focus on procedures seemed to limit their awareness of learning, it did appear to facilitate their attempt to give and receive scaffolding. There was little awareness of the self as learner, and learning as a concept was largely associated with the teacher. In their second year, children's perceptions of learning reflected changes in classroom dynamics and teaching styles and a greater awareness of learning content and strategies. The findings suggested that children look for patterns in school and construct scripts for school learning which reflect the changing dynamics of classrooms, and that peer interactions provide opportunities for developing relationships which support emergent forms of scaffolding, including metacognitive assistance to peers. Results suggested that teaching practices need to support the construction of scripts for learning and not simply scripts for routines and procedures. (Contains 25 references.) (KDFB)

We study the time taken by a language learner to correctly identify the meaning of all words in a lexicon under conditions where many plausible meanings can be inferred whenever a word is uttered. We show that the most basic form of cross-situational learning - whereby information from multiple episodes is combined to eliminate incorrect meanings - can perform badly when words are learned independently and meanings are drawn from a nonuniform distribution. If learners further assume that no two words share a common meaning, we find a phase transition between a maximally-efficient learning regime, where the learning time is reduced to the shortest it can possibly be, and a partially-efficient regime where incorrect candidate meanings for words persist at late times. We obtain exact results for the word-learning process through an equivalence to a statistical mechanical problem of enumerating loops in the space of word-meaning mappings.

We describe the application of tools from statistical mechanics to analyse the dynamics of various classes of supervised learning rules in perceptrons. The character of this paper is mostly that of a cross between a biased non-encyclopedic review and lecture notes: we try to present a coherent and self-contained picture of the basics of this field, to explain the ideas and tricks, to show how the predictions of the theory compare with (simulation) experiments, and to bring together scattered results. Technical details are given explicitly in an appendix. In order to avoid distraction we concentrate the references in a final section. In addition this paper contains some new results: (i) explicit solutions of the macroscopic equations that describe the error evolution for on-line and batch learning rules, (ii) an analysis of the dynamics of arbitrary macroscopic observables (for complete and incomplete trainingsets), leading to a general Fokker-Planck equation, and (iii) the macroscopic laws describing batch learning with complete training sets. We close the paper with a preliminary expose of ongoing research on the dynamics of learning for the case where the training set is incomplete (i.e. where the number of examples scales linearly with the network size).

This study investigates the effectiveness of the live interactive broadcast method of delivering engineering content into a standard university classroom. The Dynamics Concepts Inventory, a standardized quantitative assessment, has been administered as a pretest and final assessment for the broadcast section of dynamics for four semesters. Student attitudes toward the non-standard learning environment were assessed through an instructor-developed survey, third party interviews, and anecdotal evidence. Results of the DCI indicate that outcomes from the broadcast section are similar to national average for lecture style classes. Attitude surveys revealed a persistent sense of student isolation and frustration with lack of personal contact with the instructor, but little difficulty with the broadcast and technology assisted class room environment.

Learning and memory are acquired through long-lasting changes in synapses. In the simplest models, such synaptic potentiation typically leads to runaway excitation, but in reality there must exist processes that robustly preserve overall stability of the neural system dynamics. How is this accomplished? Various approaches to this basic question have been considered. Here we propose a particularly compelling and natural mechanism for preserving stability of learning neural systems. This mechanism is based on the global processes by which metabolic resources are distributed to the neurons by glial cells. Specifically, we introduce and study a model comprised of two interacting networks: a model neural network interconnected by synapses which undergo spike-timing dependent plasticity (STDP); and a model glial network interconnected by gap junctions which diffusively transport metabolic resources among the glia and, ultimately, to neural synapses where they are consumed. Our main result is that the biophysical constraints imposed by diffusive transport of metabolic resources through the glial network can prevent runaway growth of synaptic strength, both during ongoing activity and during learning. Our findings suggest a previously unappreciated role for glial transport of metabolites in the feedback control stabilization of neural network dynamics during learning.

The study will examine the learning rates and discount factors of the semantic and affective representations of valence during reversal learning with two states.

The study will examine the learning rates and discount factors of the semantic and affective representations of valence during reversal learning with two states.

256 p. 23 cm

256 p. 23 cm

256 p. 23 cm

Research suggests that working memory has an important role in instrumental learning when reinforcement histories of multiple options must be tracked (Collins et al., 2017; Collins & Frank, 2012, 2018). Working memory capacity not only reflects the ability to maintain items, but also to update and shield items against interference in a context-dependent manner; functions conceivably also essential to instrumental learning in an ever-changing environment. To address the relationship of working memory capacity and the dynamics of instrumental learning in a volatile environment in a previous unpublished study, we studied choice behavior and EEG of healthy young people who performed a probabilistic reversal learning task with sudden changes in reward probability (20, 50, 80%)(Humann et al., 2020). Working memory capacity, as assessed with the automated Operation Span (AOSPAN) task, positively correlated with learning performance. A hybrid Rescorla-Wagner/Pearce-Hall model was implemented to capture dynamic changes in people’s learning rate around changes in reward probability. Low-capacity participants modulated learning rates less dynamically around such changes. Their choices were more stochastic and less guided by learnt values, resulting in less stable performance and higher susceptibility to misleading probabilistic feedback. Single-trial model-based EEG analysis furthermore revealed that reward prediction errors (RPE) and learning rates were less strongly represented in cortical activity of low-capacity participants, while the centroparietal positivity, a general correlate of behavioural adaptation, was independent of working memory capacity. Whether working memory capacity modulated learning performance in this task (causation), or performance on both tasks relied on a common underlying mechanism (correlation), is an open question that we aim to address in this online behavioural study. By experimentally manipulating the load on working memory in the original task, its causal role in learning performance and dynamics can be investigated. Our second aim is to conceptually replicate our previous correlations between working memory capacity and learning performance as well as dynamic learning rate, under the load manipulation that is most similar to the load in the original task.

This paper explores one practitioner's learning development work with PhD students in a changing university context in which managerialism and financial stringency have combined. It questions how learning development practitioners can maintain their professional goals while negotiating issues arising from managerialism, financial stringency, task-oriented budgeting and inter-professional co-operation. While learning development workers have traditionally espoused a broad, colleagual view of student learning and practice which looks beyond remediation, a cash-starved sector may look for simpler, task-oriented definitions in which work is discretely allocated to different professional groups rather than dealing with dynamic, complex and overlapping processes. Hence, two teaching strategies are explored: the use of pre-course learning development tutorials as a means to tackle timeliness and efficacy of courses, and learning development contributions to supervision development courses.

Although mechanisms underlying decisions made for ourselves are well-explored, predictions of other’s choices remain relatively understudied. Recent evidence indicates that decisions for the self and another person may be performed by the same underlying cognitive mechanisms (Bavard et al., 2023), although this seems to depend on factors such as the perceived self-other similarity (Stuchlý et al., 2024). On the brain dynamics level, studies on mentalising and perspective-taking give an inconsistent picture, identifying both similarities (Jenkins et al., 2008) and differences (Tamir & Mitchell, 2010) when representing oneself vs. another person. The few studies which investigated the brain signal of predicting decisions more specifically (e.g., Harris et al., 2018) are, likewise, not conclusive or comprehensive, as they have only used vaguely described hypothetical others, making it impossible to quantify aspects such as self-other similarity or to determine the accuracy of the predicted decisions. The aim of the current study is to fill this gap by identifying and comparing the neural underpinnings of making decisions for oneself and when predicting decisions of another agent in a more information-rich environment. Participants will first make risky choices for themselves; then observe the choices of another (artificial) agent and learn their risk preference; and finally predict how this agent would decide on the same set of options that participants completed for themselves. Importantly, participants will observe and predict for one similar and one dissimilar agent, where we will precisely adjust the self-other risk preference similarity based on participants’ own risk preference. Most important is the question of whether we find evidence of the same brain signal encoding decisions for oneself and the decisions predicted for another agent (similar, dissimilar or both).

The study will examine the learning rates and discount factors of the semantic and affective representations of valence during reversal learning with two states.

The beliefs of job seekers about their own future prospects are important, but despite being a long-standing topic of interest, relatively little is known about how unemployed job seekers themselves think their job-finding probability will evolve if they remain unemployed for longer. In this study, we conduct a survey to explore this question.

We study the dynamics of supervised learning in layered neural networks, in the regime where the size $p$ of the training set is proportional to the number $N$ of inputs. Here the local fields are no longer described by Gaussian probability distributions. We show how dynamical replica theory can be used to predict the evolution of macroscopic observables, including the relevant performance measures, incorporating the old formalism in the limit $\alpha=p/N\to\infty$ as a special case. For simplicity we restrict ourselves to single-layer networks and realizable tasks.

Large-scale recurrent networks have drawn increasing attention recently because of their capabilities in modeling a large variety of real-world phenomena and physical mechanisms. This paper studies how to identify all authentic connections and estimate system parameters of a recurrent network, given a sequence of node observations. This task becomes extremely challenging in modern network applications, because the available observations are usually very noisy and limited, and the associated dynamical system is strongly nonlinear. By formulating the problem as multivariate sparse sigmoidal regression, we develop simple-to-implement network learning algorithms, with rigorous convergence guarantee in theory, for a variety of sparsity-promoting penalty forms. A quantile variant of progressive recurrent network screening is proposed for efficient computation and allows for direct cardinality control of network topology in estimation. Moreover, we investigate recurrent network stability conditions in Lyapunov's sense, and integrate such stability constraints into sparse network learning. Experiments show excellent performance of the proposed algorithms in network topology identification and forecasting.

Large-scale recurrent networks have drawn increasing attention recently because of their capabilities in modeling a large variety of real-world phenomena and physical mechanisms. This paper studies how to identify all authentic connections and estimate system parameters of a recurrent network, given a sequence of node observations. This task becomes extremely challenging in modern network applications, because the available observations are usually very noisy and limited, and the associated dynamical system is strongly nonlinear. By formulating the problem as multivariate sparse sigmoidal regression, we develop simple-to-implement network learning algorithms, with rigorous convergence guarantee in theory, for a variety of sparsity-promoting penalty forms. A quantile variant of progressive recurrent network screening is proposed for efficient computation and allows for direct cardinality control of network topology in estimation. Moreover, we investigate recurrent network stability conditions in Lyapunov's sense, and integrate such stability constraints into sparse network learning. Experiments show excellent performance of the proposed algorithms in network topology identification and forecasting.

The study will examine the discount factors of the semantic and affective representations of valence during reversal learning with two states.

The study will examine the discount factors of the semantic and affective representations of valence during reversal learning with two states.

The haptic paddle is a force-feedback joystick used at several universities in teaching System Dynamics, a core mechanical engineering undergraduate course where students learn to model dynamic systems in several domains. A second goal of the haptic paddle is to increase the accessibility of robotics and haptics by providing a low-cost device for middle and high school teachers and students, a goal which has been hindered to date by the lack of low-cost electronic solutions for motor control and computer interfacing. Prior assessments of the learning enabled by the paddles at the college level have been qualitative, consisting of anecdotal case studies illustrating student and educator belief that they enhance learning and increase student enthusiasm. In this paper, we describe haptic paddle design enhancements and provide a formal assessment of student learning when interacting with haptic paddles. Design enhancements seek to enable broad dissemination and improve the student experience by making the paddle less expensive (less than $100 for a complete system including electronics) and easier to use. Our formal assessment quantifies if and when learning occurs in a System Dynamics course featuring haptic paddle laboratories, via multiple choice quizzes presented at various time points to measure learning in the lecture, in the lab, during lab report writing, and at the end of the semester.

We present an unsupervised representation learning approach that compactly encodes the motion dependencies in videos. Given a pair of images from a video clip, our framework learns to predict the long-term 3D motions. To reduce the complexity of the learning framework, we propose to describe the motion as a sequence of atomic 3D flows computed with RGB-D modality. We use a Recurrent Neural Network based Encoder-Decoder framework to predict these sequences of flows. We argue that in order for the decoder to reconstruct these sequences, the encoder must learn a robust video representation that captures long-term motion dependencies and spatial-temporal relations. We demonstrate the effectiveness of our learned temporal representations on activity classification across multiple modalities and datasets such as NTU RGB+D and MSR Daily Activity 3D. Our framework is generic to any input modality, i.e., RGB, Depth, and RGB-D videos.

A formalism for describing the dynamics of Genetic Algorithms (GAs) using methods from statistical mechanics is applied to the problem of generalization in a perceptron with binary weights. The dynamics are solved for the case where a new batch of training patterns is presented to each population member each generation, which considerably simplifies the calculation. The theory is shown to agree closely to simulations of a real GA averaged over many runs, accurately predicting the mean best solution found. For weak selection and large problem size the difference equations describing the dynamics can be expressed analytically and we find that the effects of noise due to the finite size of each training batch can be removed by increasing the population size appropriately. If this population resizing is used, one can deduce the most computationally efficient size of training batch each generation. For independent patterns this choice also gives the minimum total number of training patterns used. Although using independent patterns is a very inefficient use of training patterns in general, this work may also prove useful for determining the optimum batch size in the case where patterns are recycled.

A formalism for describing the dynamics of Genetic Algorithms (GAs) using methods from statistical mechanics is applied to the problem of generalization in a perceptron with binary weights. The dynamics are solved for the case where a new batch of training patterns is presented to each population member each generation, which considerably simplifies the calculation. The theory is shown to agree closely to simulations of a real GA averaged over many runs, accurately predicting the mean best solution found. For weak selection and large problem size the difference equations describing the dynamics can be expressed analytically and we find that the effects of noise due to the finite size of each training batch can be removed by increasing the population size appropriately. If this population resizing is used, one can deduce the most computationally efficient size of training batch each generation. For independent patterns this choice also gives the minimum total number of training patterns used. Although using independent patterns is a very inefficient use of training patterns in general, this work may also prove useful for determining the optimum batch size in the case where patterns are recycled.

A formalism for describing the dynamics of Genetic Algorithms (GAs) using methods from statistical mechanics is applied to the problem of generalization in a perceptron with binary weights. The dynamics are solved for the case where a new batch of training patterns is presented to each population member each generation, which considerably simplifies the calculation. The theory is shown to agree closely to simulations of a real GA averaged over many runs, accurately predicting the mean best solution found. For weak selection and large problem size the difference equations describing the dynamics can be expressed analytically and we find that the effects of noise due to the finite size of each training batch can be removed by increasing the population size appropriately. If this population resizing is used, one can deduce the most computationally efficient size of training batch each generation. For independent patterns this choice also gives the minimum total number of training patterns used. Although using independent patterns is a very inefficient use of training patterns in general, this work may also prove useful for determining the optimum batch size in the case where patterns are recycled.

This collection of abstracts is part of a continuing series providing information on recent doctoral dissertations. The 47 titles deal with a variety of topics, including the following: (1) the effect of source variation upon children's reactions to television commercials; (2) covert observation as a means of measuring the affective states of television viewers; (3) viewers' responses to a film as a socially situated event; (4) children's understanding of television reality; (5) media learning and adolescent developmental tasks; (6) the dynamics of communication and information in groups; (7) an information processing approach to uses and gratifications; (8) privacy and media encroachment; (9) the changing images of females and males in television commercials; (10) commercial broadcasters' perceptions of entry-level employment requirements for college graduates in the broadcast industry; (11) transnational advertising and the mass media in Latin America; (12) longitudinal patterns of television viewing and adolescent role socialization; (13) the relationship between persistence, attention, and observational learning of television's program content; (14) the film industry and the Vietnam war; (15) the effect of televised violence; (16) the use of videotape for studying nonverbal behavior in social settings; (17) United States Supreme Court interpretations of "public interest" in decisions concerning the media from 1927 to 1979. (FL)

We introduce a theory of sequential causal inference in which learners in a chain estimate a structural model from their upstream teacher and then pass samples from the model to their downstream student. It extends the population dynamics of genetic drift, recasting Kimura's selectively neutral theory as a special case of a generalized drift process using structured populations with memory. We examine the diffusion and fixation properties of several drift processes and propose applications to learning, inference, and evolution. We also demonstrate how the organization of drift process space controls fidelity, facilitates innovations, and leads to information loss in sequential learning with and without memory.

Human culture is widely believed to undergo evolution, via mechanisms rooted in the nature of human cognition. A number of theories predict the kinds of human learning strategies, as well as the population dynamics that result from their action. There is little work, however, that quantitatively examines the evidence for these strategies and resulting cultural evolution within human populations. One of the obstacles is the lack of individual-level data with which to link transmission events to larger cultural dynamics. Here, we address this problem with a rich quantitative database from the East Asian board game known as Go. We draw from a large archive of Go games spanning the last six decades of professional play, and find evidence that the evolutionary dynamics of particular cultural variants are driven by a mix of individual and social learning processes. Particular players vary dramatically in their sensitivity to population knowledge, which also varies by age and nationality. The dynamic patterns of opening Go moves are consistent with an ancient, ongoing arms race within the game itself.

Using the cavity method and diagrammatic methods, we model the dynamics of batch learning of restricted sets of examples, widely applicable to general learning cost functions, and fully taking into account the temporal correlations introduced by the recycling of the examples.

Includes bibliographical references

Transportation has been considered as the backbone of the economy for the past many years. Unfortunately, since few years due to the uncontrolled urbanization and inadequate planning, countries are facing problem of congestion. The congestion is hindering the economic growth and also causing environmental issues. This has caused serious concerns among the major economies of the world, especially in Asia-Pacific region. Many countries are playing an active role in eradicating this problem and some have been quite successful so far. Malaysia, being a major ASEAN economy is also tackling with this huge problem. The authorities are committed to solve the issue. In this regard, solving the issue leveraging the use of big data analytics has become crucial. The authorities can form a complete robust framework based on big data analytics and decision making process to solve the issue effectively. The work focuses and observes the traffic data samples and analyzes the accuracy of machine learning algorithms, which helps in decision making. Yet, here is a lot to be done if the government needs to solve the problem effectively. Supposedly, a comprehensive big data transport framework leveraging machine learning, is one way to solve the issue. 

The purpose of the study was to model the processes involved in the development of creativity and discuss how motivation and teaching strategies figure into a creativity-supportive learning environment. The subjects included 28 college students enrolled in an Introduction to Instructional Media course. A questionnaire and interviews were used for the collection of data. Over the course of the semester, the students were tasked with making a "creative multimedia video website". The results of the study reveal that the motivation of students in areas such as task value and self-efficacy were enhanced by this exercise. Moreover, this study developed effective strategies for the promotion of creativity and the enhancement of high-level meta-cognitive processes. (Contains 4 tables and 1 figure.)

In this work we analyze the solutions of a simple system of coupled phase oscillators in which the connectivity is learned dynamically. The model is inspired in the process of learning of birdsong by oscine birds. An oscillator acts as the generator of a basic rhythm, and drives slave oscillators which are responsible for different motor actions. The driving signal arrives to each driven oscillator through two different pathways. One of them is a "direct" pathway. The other one is a "reinforcement" pathway, through which the signal arrives delayed. The coupling coefficients between the driving oscillator and the slave ones evolve in time following a Hebbian-like rule. We discuss the conditions under which a driven oscillator is capable of learning to lock to the driver. The resulting phase difference and connectivity is a function of the delay of the reinforcement. Around some specific delays, the system is capable to generate dramatic changes in the phase difference between the driver and the driven systems. We discuss the dynamical mechanism responsible for this effect, and possible applications of this learning scheme.

In this work we analyze the solutions of a simple system of coupled phase oscillators in which the connectivity is learned dynamically. The model is inspired in the process of learning of birdsong by oscine birds. An oscillator acts as the generator of a basic rhythm, and drives slave oscillators which are responsible for different motor actions. The driving signal arrives to each driven oscillator through two different pathways. One of them is a "direct" pathway. The other one is a "reinforcement" pathway, through which the signal arrives delayed. The coupling coefficients between the driving oscillator and the slave ones evolve in time following a Hebbian-like rule. We discuss the conditions under which a driven oscillator is capable of learning to lock to the driver. The resulting phase difference and connectivity is a function of the delay of the reinforcement. Around some specific delays, the system is capable to generate dramatic changes in the phase difference between the driver and the driven systems. We discuss the dynamical mechanism responsible for this effect, and possible applications of this learning scheme.

In this work we analyze the solutions of a simple system of coupled phase oscillators in which the connectivity is learned dynamically. The model is inspired in the process of learning of birdsong by oscine birds. An oscillator acts as the generator of a basic rhythm, and drives slave oscillators which are responsible for different motor actions. The driving signal arrives to each driven oscillator through two different pathways. One of them is a "direct" pathway. The other one is a "reinforcement" pathway, through which the signal arrives delayed. The coupling coefficients between the driving oscillator and the slave ones evolve in time following a Hebbian-like rule. We discuss the conditions under which a driven oscillator is capable of learning to lock to the driver. The resulting phase difference and connectivity is a function of the delay of the reinforcement. Around some specific delays, the system is capable to generate dramatic changes in the phase difference between the driver and the driven systems. We discuss the dynamical mechanism responsible for this effect, and possible applications of this learning scheme.

Two papers in this report deal with early childhood. "Developmental Factors--Early Cognitive Growth" discusses the experimental-analytical and existential-experimental approaches to religious cognitive growth. The second paper, "Developmental Factors--Early Emotional-Social Growth" asserts that emotional and social integrity has to be achieved through progressive but gradual maturation of the personality from infancy onward. (MS)

Learning would be a convincing method to achieve coordination on an equilibrium. But does learning converge, and to what? We answer this question in generic 2-player, 2-strategy games, using Experience-Weighted Attraction (EWA), which encompasses many extensively studied learning algorithms. We exhaustively characterize the parameter space of EWA learning, for any payoff matrix, and we understand the generic properties that imply convergent or non-convergent behaviour in 2 x 2 games. Irrational choice and lack of incentives imply convergence to a mixed strategy in the centre of the strategy simplex, possibly far from the Nash Equilibrium (NE). In the opposite limit, in which the players quickly modify their strategies, the behaviour depends on the payoff matrix: (i) a strong discrepancy between the pure strategies describes dominance-solvable games, which show convergence to a unique fixed point close to the NE; (ii) a preference towards profiles of strategies along the main diagonal describes coordination games, with multiple stable fixed points corresponding to the NE; (iii) a cycle of best responses defines discoordination games, which commonly yield limit cycles or low-dimensional chaos. While it is well known that mixed strategy equilibria may be unstable, our approach is novel from several perspectives: we fully analyse EWA and provide explicit thresholds that define the onset of instability; we find an emerging taxonomy of the learning dynamics, without focusing on specific classes of games ex-ante; we show that chaos can occur even in the simplest games; we make a precise theoretical prediction that can be tested against data on experimental learning of discoordination games.

The population of the hermit crab Pagurus samuelis at Carmel River Beach, Carmel, California, was surveyed and the results were statistically treated to determine the distribution of the species. Groups within the population were observed to evaluate the stability of the groups. The population was found to be contagiously distributed with varying degrees of stability exhibited by the groups. The factors governing stability were the restriction imposed by the microenvironment and the amount of time during which the area containing the crabs was flooded. Laboratory and field observations of individual behavior showed the species to be of an aggressive nature. The learning capacity of P. samuelis was the subject of an experiment in which the performance of individual crabs in a simple T-maze was evaluated; the experiment showed that the crabs were capable of improving their performance with time. (Author)

This project developed fundamental theory and novel algorithms for adaptive learning in autonomous collective-agent systems. The first goal was to develop a new mathematical framework for analyzing the dynamical mechanisms that support learning in a range of novel information processing substrates. The second goal was to adapt this single-agent learning theory and associated learning algorithms to the distributed setting in which a population of autonomous agents communicate to achieve a desired group task. The results provide a mathematically sound basis for quantitatively measuring 1) the degree of individual-agent intelligence; 2) the significance of agent-environment interaction; and 3) the emergence of cooperation in agent collectives. A novel feature was that the theoretical approach synthesized recent results in the areas of machine learning, statistical inference, statistical mechanics, nonlinear dynamics, and pattern formation theory and then applied them to single-agent learning and adaptive learning in agent collectives. The central impact of the project will be its systematic and quantitative approach to predicting behavioral complexity in agents and in their environments. This will provide a foundation for measuring agent and agent-collective intelligence which, in turn, should allow for systematic engineering and monitoring of these systems.

Estimation of the operational topology of the power grid is necessary for optimal market settlement and reliable dynamic operation of the grid. This paper presents a novel framework for topology estimation for general power grids (loopy or radial) using time-series measurements of nodal voltage phase angles that arise from the swing dynamics. Our learning framework utilizes multivariate Wiener filtering to unravel the interaction between fluctuations in voltage angles at different nodes and identifies operational edges by considering the phase response of the elements of the multivariate Wiener filter. The performance of our learning framework is demonstrated through simulations on standard IEEE test cases.

People are increasingly relying on the Web and social media to find solutions to their problems in a wide range of domains. In this online setting, closely related problems often lead to the same characteristic learning pattern, in which people sharing these problems visit related pieces of information, perform almost identical queries or, more generally, take a series of similar actions. In this paper, we introduce a novel modeling framework for clustering continuous-time grouped streaming data, the hierarchical Dirichlet Hawkes process (HDHP), which allows us to automatically uncover a wide variety of learning patterns from detailed traces of learning activity. Our model allows for efficient inference, scaling to millions of actions taken by thousands of users. Experiments on real data gathered from Stack Overflow reveal that our framework can recover meaningful learning patterns in terms of both content and temporal dynamics, as well as accurately track users' interests and goals over time.