Browsing by Subject "learning"

Now showing 1 - 14 of 14

The Cognitive Consequences of Childhood Adversity: An Investigation of Feedback Learning Strategies and the Sensitized-Specialization Hypothesis
(2019-07) Young, Ethan
Stressful environments have a profound impact on children. The prevailing view is that adverse experiences in childhood impairs the mind and derails development. In contrast, the current research draws on the specialization hypothesis, which proposes that children should develop specialized cognitive abilities that are adapted to adverse environments. This view focuses on the strengths of people who have experienced adversity instead of exclusively on their weaknesses. In this dissertation, I test how different learning abilities might be enhanced by exposure to early adversity. I conducted four experimental tests of this hypothesis in relation to reversal learning performance, examining how growing up in unpredictable versus predictable environments are associated with different learning strategies. I hypothesize that growing up in a more predictable environment should be associated with the use of learning strategies that integrate information over longer periods of time, whereas growing up in a more unpredictable environment should be associated with learning strategies that rely on recent information. Furthermore, based on previous studies, I hypothesize that such learning strategies should be activated by the threat of uncertainty in the current environment. Across 4 experiments, I tested how exposure to childhood unpredictability impacts both overall reversal learning performance and trial-by-trial learning styles and tested whether current, experimentally manipulated cues of economic uncertainty modulated reversal learning outcomes. Findings were mixed and inconsistent. On average, overall indicators of reversal learning performance seem to be impaired by exposure to greater childhood unpredictability. For trial-by-trial learning performance, there were some main effects of childhood unpredictability, but they were inconsistent across experiments. Finally, for all outcomes and experiments, the current context did not moderate the effect of early childhood. I discuss possible explanations for these inconsistent findings and lay out new avenues for future research. Despite these inconsistent findings, this research remains important because it could help to identify the types of learning strategies that are most effective for success among people from disadvantaged backgrounds.
The Conceptual Framework for the Professional Education Programs in the College of Education and Human Development: University of Minnesota
(University of Minnesota, 2005-08) University of Minnesota: College of Education and Human Development
The primary purposes of the College of Education and Human Development (CEHD) at the University of Minnesota are to advance knowledge in the field of education, to prepare personnel for educational and human development positions, and to provide leadership to educational and human development agencies. The college intends to continue to build on its national reputation in the area of teacher preparation.
For the Love of Technology: How Aesthetics Define Emotions in a Digital Education Setting
(2021) Brower, Autumn
Educators often admit that they are aware that emotion plays a significant role in students’ educational success, yet most of the scientific literature measures cognition in education to the exclusion of emotion. This study is intended to be a proof of concept design for future research. Its goal is to assess how an individual’s aesthetic value of a product might be a way to gauge emotion in educational settings. Three faculty members at the University of Minnesota were interviewed about their viewpoints pertaining to the product design of a static Canvas page and asked to evaluate its design based on its visceral, behavioral, and reflective beauty. Page orientation and font were used to represent product design. Results of the interviews showed that readability was the most frequently mentioned reason people are drawn to certain aesthetic features of a product’s design in digital education, followed by alignment, accessibility, mobile devices, tradition, and font personalities. Additionally, this paper evaluates the participants’ valence response; their responses to the design’s functionality; and their thoughts on meaningfulness as they relate to Norman’s (2007) three aesthetic levels of product design. At the end of the paper, suggestions for how we might use this data to increase productivity in our classes and enhance educational technology are addressed. Future directions for how these results might apply to cognition, emotion, and computation are also discussed.
A Game-Based Solution to the Lack of Training and Assessment Opportunities for Spatial Reasoning
(2023-01) VanMeerten, Nicolaas
Spatial reasoning is an important skill that people use on a daily basis. There is also strong evidence that people with enhanced spatial reasoning skills are more likely to pursue successful careers related to Science, Technology, Engineering, and Mathematics (STEM). Spatial reasoning skills are also malleable, which suggests that spatial reasoning training and assessment could be used to enhance academic outcomes in STEM. However, there are relatively few readily accessible training or assessment opportunities for spatial reasoning. Commercial video games should be adapted to create more spatial reasoning training environments. Video games provide unique affordances that support training and learning, including: (1) delivering the appropriate level of challenge and (2) the ease of assessment integration. I found evidence that there is a relationship between performance in Optica, a mobile-puzzle game, and spatial reasoning skills among middle-school students. Specifically, I discovered a relationship between the number of levels completed in Optica and score on the PSVT: R by comparing multiple linear regression models with Akaike Information Criteria. Thus, Optica has shown potential as a suitable virtual environment for training and assessing spatial reasoning skills. Although there were limitations to this study, they can be remedied by updates to the design of the game, telemetry collection, and enhanced experimental design. I believe that Optica should be iterated upon to develop it into a fully-fledged game environment for training and assessing spatial reasoning skills, which will benefit many areas of STEM simultaneously.
Integration of Emerging Learning Technologies in Secondary Schools : A Burkina Faso Case Study
(2017-01) Zongo, Romaric
The purpose of this dissertation was to document the perspectives and attitudes of secondary education teachers and administrators about the perceived benefits and challenges of integrating new Emerging Learning Technologies (ELTs) in the classroom. Education has become one of the biggest challenges in the African nation of Burkina Faso where teachers are routinely confronted with material shortages, lack of curriculum, lack of equipment, and lack of opportunity for self-conducted learning. To overcome these challenges, educators are using Emerging Learning Technologies (ELTs) to help improve the quality of teaching and to increase student access to these learning opportunities. This study examined three core questions that specifically focused is on the ways in which ELTs are perceived as different from previously used technologies in Burkina Faso (i.e., radio and television). 1. What are the perspectives of secondary level educators and administrators regarding the use of ELTs in Burkina Faso? 2. In the local educational contexts of cities and rural areas, how do educators and administrators experience the use of ELTs in education? 3. What are the benefits and challenges of using ELTs for educational purposes in Burkina Faso? Study findings indicated that the pedagogical use of ELTs in secondary education was not contributing to teaching and learning in secondary schools at this time. Analysis of the collected data found that the added value of the use of ELTs in education depended mainly on their daily adaption by students, teachers, and administrative staff. However, the use of ELTs in secondary education in Burkina Faso is infrequent and not widely embraced by school administrators and teachers. Future adoption of ELTs may someday impact educational outcomes but it will take more than top-down political directives to achieve this outcome.
Learning of Unknown Environments in Goal-Directed Guidance and Navigation Tasks: Autonomous Systems and Humans
(2017-12) Verma, Abhishek
Guidance and navigation in unknown environments requires learning of the task environment simultaneous to path planning. Autonomous guidance in unknown environments requires a real-time integration of environment sensing, mapping, planning, trajectory generation, and tracking. For brute force optimal control, the spatial environment should be mapped accurately. The real-world environments are in general cluttered, complex, unknown, and uncertain. An accurate model of such environments requires to store an enormous amount of information and then that information has to be processed in optimal control formulation, which is not computationally cheap and efficient for online operations of autonomous guidance systems. On the contrary, humans and animals are in general able to navigate efficiently in unknown, complex, and cluttered environments. Like autonomous guidance systems, humans and animals also do not have unlimited information processing and sensing capacities due to their biological and physical constraints. Therefore, it is relevant to understand cognitive mechanisms that help humans learn and navigate efficiently in unknown environments. Such understanding can help to design planning algorithms that are computationally efficient as well as better understand how to improve human-machine interfaces in particular between operators and autonomous agents. This dissertation is organized in three parts: 1) computational investigation of environment learning in guidance and navigation (chapters 3 and 4), 2) investigation of human environment learning in guidance tasks (chapters 5 and 6), and 3) autonomous guidance framework based on a graph representation of environment using subgoals that are invariants in agent-environment interactions (chapter 7). In the first part, the dissertation presents a computational framework for learning autonomous guidance behavior in unknown or partially known environments. The learning framework uses a receding horizon trajectory optimization associated with a spatial value function (SVF). The SVF describes optimal (e.g. minimum time) guidance behavior represented as cost and velocity at any point in geographical space to reach a specified goal state. For guidance in unknown environments, a local SVF based on current vehicle state is updated online using environment data from onboard exteroceptive sensors. The proposed learning framework has the advantage in that it learns information directly relevant to the optimal guidance and control behavior enabling optimal trajectory planning in unknown or partially known environments. The learning framework is evaluated by measuring performance over successive runs in a 3-D indoor flight simulation. The test vehicle in the simulations is a Blade-Cx2 coaxial miniature helicopter. The environment is a priori unknown to the learning system. The dissertation investigates changes in performance, dynamic behavior, SVF, and control behavior in body frame, as a result of learning over successive runs. In the second part, the dissertation focuses on modeling and evaluating how a human operator learns an unknown task environment in goal-directed navigation tasks. Previous studies have showed that human pilots organize their guidance and perceptual behavior using the interaction patterns (IPs), i.e., invariants in their sensory-motor processes in interactions with the task space. However, previous studies were performed in known environments. In this dissertation, the concept of IPs is used to build a modeling and analysis framework to investigate human environment learning and decision-making in navigation of unknown environments. This approach emphasizes the agent dynamics (e.g., a vehicle controlled by a human operator), which is not typical in simultaneous navigation and environment learning studies. The framework is applied to analyze human data from simulated first-person guidance experiments in an obstacle field. Subjects were asked to perform multiple trials and find minimum-time routes between prespecified start and goal locations without priori knowledge of the environment. They used a joystick to control flight behavior and navigate in the environment. In the third part, the subgoal graph framework used to model and evaluate humans is extended to an autonomous guidance algorithm for navigation in unknown environments. The autonomous guidance framework based on subgoal graph is an improvement to the SVF based guidance and learning framework presented in the first part. The latter uses a grid representation of the environment, which is computationally costly in comparison to the graph based guidance model.
Leveraging Sparsity and Low Rank for Large-Scale Networks and Data Science
(2015-05) Mardani, Morteza
We live in an era of ``data deluge," with pervasive sensors collecting massive amounts of information on every bit of our lives, churning out enormous streams of raw data in a wide variety of formats. While big data may bring ``big blessings," there are formidable challenges in dealing with large-scale datasets. The sheer volume of data makes it often impossible to run analytics using central processors and storage units. Network data are also often geographically spread, and collecting the data might be infeasible due to communication costs or privacy concerns. Disparate origin of data also makes the datasets often incomplete, and thus a sizable portion of entries are missing. Moreover, large-scale data are prone to contain corrupted measurements, communication errors, and even su ffer from anomalies due to cyberattacks. Moreover, as many sources continuously generate data in real time, analytics must often be performed online as well as without an opportunity to revisit past data. Last but not least, due to variety, data is typically indexed by multiple dimensions. Towards our vision to facilitate learning, this thesis contributes to cope with these challenges via leveraging the low intrinsic-dimensionality of data by means of sparsity and low rank. To build a versatile model capturing various data irregularities, the present thesis focuses first on a low-rank plus compressed-sparse matrix model, which proves successful in unveiling trffia c anomalies in backbone networks. Leveraging the nuclear and \ell_1-norm, exact reconstruction guarantees are established for a convex estimator of the unknowns. Inspired by the crucial task of network tra ffic monitoring, the scope of this model and recovery task is broaden to a tomographic task of jointly mapping out nominal and anomalous tra ffic from undersampled linear measurements. Despite the success of nuclear-norm minimization in capturing the data low dimensionality, it scales very poorly with the data size mainly due to its tangled nature. This indeed hinders decentralized and streaming analytics. To mitigate this computational challenge, this thesis puts forth a neat framework which permeates benefits from a bilinear characterization of nuclear-norm to bring separability at the expense of nonconvexity. Notwithstanding, it is proven that under certain conditions stationary points of nonconvex program coincide with the optimum of the convex counterpart. Using this idea along with theory of alternating minimization we develop lightweight algorithms with low communication-overhead for in-network processing; and provably convergent online ones suitable for streaming analytics. All in all, the major innovative claim is that even with the budget of distributed computation and sequential acquisition one can hope to achieve accurate reconstruction guarantees o ffered by the batch nuclear-norm minimization. Finally, inspired by the k-space data interpolation task appearing in dynamic magnetic resonance imaging, a novel tensor subspace learning framework is introduced to handle streaming multidimensional data. It capitalizes on the PARAFAC decomposition and e effects low tensor rank by means of the Tykhonov regularization, that enjoys separability and offers real-time MRI reconstruction tailoring e.g., image-guided radiation therapy applications.
Measuring the Role of Vulnerability in the Classroom
(2020-01) Hennessey, Ryan
Education has been diverging from a human- to a digital-centered practice with increasing technological advances and initiatives at school districts put tablets or computers in the hands of every student. It is crucial, now more than ever, to provide engagement that will actively rehumanize education. (Brown, 2012). Re-introducing vulnerability has the potential to spark creativity, foster innovation, and create meaningful change in our increasingly disconnected classrooms. (Brown, 2010). This thesis investigates the potential of vulnerability to inform pedagogical practices and amplify human connections made in the classroom. It presents an action research study conducted in my own middle school classroom. I provide details on the growth around my experiences teaching, modeling, and learning about the role of vulnerability. I explore vulnerability as the special ingredient in the classroom that enhances all other pedagogical practices, and outline a future direction for this work.
Real-Time System Identification and Control of Engine System Using Least Squares Learning and Simplex Tessellation
(2022-12) Tranquillo, Holden
To aid in engine control for achieving the stable combustion of varying cetane level fuels, a computationally efficient algorithm for the online learning of an engine model based on real-time input and output measurements is developed. Innovations in engine technology has led to the feasibility of robust, multi-fuel engine systems capable of operating on unknown or non-ideal fuel types. To attain such performance, advanced control strategies must be implemented in order to achieve stable engine combustion using such fuels. The method developed in this work, based on piecewise-linear modeling via discrete nodes and recursive linear least squares is first derived for the one-dimensional system of injection timing and combustion phasing. The learning model is then used for adaptive feedforward and feedback control of the SISO system in simulation using a gaussian process model as a virtual engine. The algorithm is then extended to the two-input/two-output system of injection timing and fuel mass and their effect on combustion phasing and indicated mean effective pressure (IMEP). Data generated using computational fluid mechanics is used to supplement experimental data in the development of the 2D model. The theory of barycentric and affine coordinates is introduced and applied to the concept of piecewise planes to approximate nonlinear surfaces. The learning model is utilized in an adaptive MIMO feedforward algorithm to control the engine to a desired combustion phasing and IMEP. Additionally, a decoupled integral feedback control scheme is presented and shown effective in simulation. A generalization of the learning algorithm for higher dimensions is made in order to model higher order systems. Specifically, simplex tessellation and barycentric coordinates as regressor coefficients are shown to generalize node locating and updating in arbitrary dimensions. The generalized learning algorithm is applied to a synthetic three-input data set in order show feasibility of the model for higher order nonlinear systems. The algorithm developed in this work is a unique, generalized, data-driven model capable of the real-time learning and control of multi-dimensional systems. The computational efficiency and generalization of the method allows for the real-time system identification of engine systems that are operating in unknown or untested environments. Existing engine models lack the efficiency to perform at the operating times seen in internal combustion engines. Implemented in a physical engine, the developed algorithm could be used for adaptive modeling of the system when undergoing a fuel or environmental change, which in turn can be used to aid in adaptive control of the engine. In commercial application, the real-time learning model could be used to decrease or eliminate the traditional in-house testing of engines required for lookup table generation, which would in turn decrease the time and cost in getting the engine to final application.
Scalable and Ensemble Learning for Big Data
(2019-05) Traganitis, Panagiotis
The turn of the decade has trademarked society and computing research with a ``data deluge.'' As the number of smart, highly accurate and Internet-capable devices increases, so does the amount of data that is generated and collected. While this sheer amount of data has the potential to enable high quality inference, and mining of information, it introduces numerous challenges in the processing and pattern analysis, since available statistical inference and machine learning approaches do not necessarily scale well with the number of data and their dimensionality. In addition to the challenges related to scalability, data gathered are often noisy, dynamic, contaminated by outliers or corrupted to specifically inhibit the inference task. Moreover, many machine learning approaches have been shown to be susceptible to adversarial attacks. At the same time, the cost of cloud and distributed computing is rapidly declining. Therefore, there is a pressing need for statistical inference and machine learning tools that are robust to attacks and scale with the volume and dimensionality of the data, by harnessing efficiently the available computational resources. This thesis is centered on analytical and algorithmic foundations that aim to enable statistical inference and data analytics from large volumes of high-dimensional data. The vision is to establish a comprehensive framework based on state-of-the-art machine learning, optimization and statistical inference tools to enable truly large-scale inference, which can tap on the available (possibly distributed) computational resources, and be resilient to adversarial attacks. The ultimate goal is to both analytically and numerically demonstrate how valuable insights from signal processing can lead to markedly improved and accelerated learning tools. To this end, the present thesis investigates two main research thrusts: i) Large-scale subspace clustering; and ii) unsupervised ensemble learning. The aforementioned research thrusts introduce novel algorithms that aim to tackle the issues of large-scale learning. The potential of the proposed algorithms is showcased by rigorous theoretical results and extensive numerical tests.
Search, Occupational Choice and Learning
(2015-08) Buyukbasaran, Tayyar
This thesis examines the labor market effects of incomplete information about workers' own job-finding process and best occupations fitting to them. Search outcomes convey information about workers' job finding abilities and appropriate occupations suited to them, and workers use this information to infer their types. This learning \ process generates endogenous heterogeneity in occupational choices and workers beliefs. Our theory explains how unemployment can affect labor market decisions including the occupational choices. Characterization results in a simple value function with reservation level of prior belief property that is similar to reservation wage property. Some interesting facts about both micro and macro data are identified and our model's explanation of these facts is discussed. In particular, our characterization gives rational for why workers with less experience in searching have (1) longer unemployment duration and (2) higher probability of changing occupation by reemployment, and (3) why shifts in Beveridge curve may be observed. Theory can also be used to (4) explain the discouraged worker phenomenon.
Sex Differences, Physiological Response, and Emotion
(2022-06) Baumann, Ashley M
Females have a higher prevalence for PTSD and other anxiety disorders than males, thus fluctuating sex hormones, such as estrogen, are considered to play a role. Research suggests that during predictable cue tasks, high estrogen females had greater startle response toward predictable tasks compared to unpredictable and control tasks. The current study used physiological responses and self-reported measures to investigate fear response during the oddball task within naturally and unnaturally cycling females. The oddball task consisted of five time points, consistent with control, unpredictable, predictable, extinction, and control blocks. Participants viewed a randomized slide show of three visual stimuli consisting of natural and control images. Participants (n = 26) were split into high or low estrogen groups and were placed in a separate group if using a hormonal contraceptive. Results found that, overall, participants had a greater positive affect at time two than at time five, F(4,80) = 3.832, p = .007. Given the small sample size, a second set of analyses assessed high estrogen level females and those using hormonal contraceptive (lower estrogen) after time the first control (time two) and after the unpredictable and predictable block (time three). Results found between group differences in state anxiety, such that HC females had greater state anxiety than the high estrogen group, F(1,12) = 4.880 , p = .047. These results were opposite for our hypotheses that overall, females with high estrogen levels will have greater self-reported mood, anxiety, and physiological response across the study. Results also opposed our hypothesis such that a group by time interaction revealed participants in the HC group had significantly higher positive affect at time two which decreased at time three, F(1,12) = 4.931, p = .046; This significant difference between time points occurred only in participants using HC.
Transforming the University: Preliminary Report of the Knowledge Management Technology Task Force
(University of Minnesota, 2006-03-27) Olson, Debra; Perkowski, Linda
The deliverable generated by the Knowledge Management Task Force is a plan for an Academic Health Center knowledge management system that is supported by technology, is continually assessed, and is delivered within a culture comprised of specific, supportive features.
When and How Does Workplace Envy Promote Job Performance? A Study on the Conditions and Mechanisms for the Functional Role of Envy in Workplace Behavior
(2014-06) Lee, Ki Yeong
In this study, I develop and test a model that explains when and why workplace envy can enhance task performance and organizational citizenship behaviors. Drawing on counterfactual theory, I propose that workplace envy plays a functional role: employees who envy coworkers learn from their envied targets via systematic information processing, especially when the enviers have high core self-evaluations (CSE) or when the envied targets provide help to enviers. To further understand the social influence of envy in triggering interpersonal dynamic processes, I delineate the processes and conditions that will prompt targets of envy to help enviers. I propose that envied targets are likely to perceive envy and will try to appease enviers by extending help, especially when enviers have central positions in friendship networks and thus can potentially undermine the target's workplace social relationships. I collect data from Korean bank tellers and insurance sales agents and use a round robin design showing that the envy-learning relationship is contingent on CSE and received help from the target and that learning from coworkers contributes to job performance via work engagement. In addition, targets are likely to perceive the envy but are not motivated to extend help even when the envier has high friendship network centrality. I discuss the implications and limitations of the study.

University Digital Conservancy

Browse by Subject

Browsing by Subject "learning"