4,692 Hits in 5.2 sec

Faster Reinforcement Learning Using Active Simulators [article]

Vikas Jain, Theja Tulabandhula
2017 arXiv   pre-print
In this work, we propose several online methods to build a learning curriculum from a given set of target-task-specific training tasks in order to speed up reinforcement learning (RL).  ...  Unlike traditional transfer learning, we consider creating a sequence from several training tasks in order to provide the most benefit in terms of reducing the total time to train.  ...  Introduction In reinforcement learning (RL), the knowledge obtained from training on a source task can be transferred to learn a target task more efficiently [27] .  ... 
arXiv:1703.07853v2 fatcat:qg5mqb2sejgr3mj7s6s64hicay

Curriculum Learning for Reinforcement Learning Domains: A Framework and Survey [article]

Sanmit Narvekar and Bei Peng and Matteo Leonetti and Jivko Sinapov and Matthew E. Taylor and Peter Stone
2020 arXiv   pre-print
Despite many advances over the past three decades, learning in many domains still requires a large amount of interaction with the environment, which can be prohibitively expensive in realistic scenarios  ...  Reinforcement learning (RL) is a popular paradigm for addressing sequential decision tasks in which the agent has only limited environmental feedback.  ...  Part of this work has taken place in the Learning Agents Research Group (LARG) at the Artificial Intelligence Laboratory, The University of Texas at Austin. LARG re-  ... 
arXiv:2003.04960v2 fatcat:iacmqeb7jjeezpo27jsnzuqb7u

Enhanced POET: Open-Ended Reinforcement Learning through Unbounded Invention of Learning Challenges and their Solutions [article]

Rui Wang, Joel Lehman, Aditya Rawal, Jiale Zhi, Yulun Li, Jeff Clune, Kenneth O. Stanley
2020 arXiv   pre-print
Importantly, both limitations pose impediments not only for POET, but for the pursuit of open-endedness in general.  ...  Creating open-ended algorithms, which generate their own never-ending stream of novel and appropriately challenging learning opportunities, could help to automate and accelerate progress in machine learning  ...  We are grateful to Leon Rosenshein, Joel Snow, Thaxton Beesley, the Colorado Data Center team and the entire Opus Team at Uber for providing our computing platform and for technical support.  ... 
arXiv:2003.08536v2 fatcat:qll5ldmbvrgyvfrnj3ydwuzqa4

Learning Purely Tactile In-Hand Manipulation with a Torque-Controlled Hand

Leon Sievers, Johannes Pitz, Berthold Bäuml
2022 arXiv   pre-print
We efficiently train in a precisely modeled and identified rigid body simulation with off-policy deep reinforcement learning, significantly sped up by a domain adapted curriculum, leading to a moderate  ...  We show that a purely tactile dextrous in-hand manipulation task with continuous regrasping, requiring permanent force closure, can be learned from scratch and executed robustly on a torque-controlled  ...  (to our knowledge such a sim2real transfer for a tactile task has never been shown before). • The training is efficient with a compute budget of CPU days instead of CPU years.  ... 
arXiv:2204.03698v2 fatcat:3lwmt5vzd5g5pm7t7mwe67qdz4

E-Learning in Dental Schools in the Times of COVID-19: A Review and Analysis of an Educational Resource in Times of the COVID-19 Pandemic

Daniel Chavarría-Bolaños, Adrián Gómez-Fernández, Carmen Dittel-Jiménez, Mauricio Montero-Aguilar
2020 Odovtos International Journal of Dental Sciences  
the virtualization environment as a teaching/learning experience.  ...  Additionally, the experience of the University of Costa Rica Faculty of Dentistry is presented, as it was evident that some of the key elements in a e-learning environment needed a quick enhancement and  ...  during the planning and implementation of the virtualization process during the COVID-19 emergency, and for all the information provided for the elaboration of this manuscript.  ... 
doi:10.15517/ijds.2020.41813 fatcat:iogpgfjuzbcb5jyqt22kkcspf4

Meta Automatic Curriculum Learning [article]

Rémy Portelas, Clément Romac, Katja Hofmann, Pierre-Yves Oudeyer
2021 arXiv   pre-print
parkour environments with learners of varying morphologies.  ...  Training on diverse tasks has been identified as a key ingredient for good generalization, which pushed researchers towards using rich procedural task generation systems controlled through complex continuous  ...  Introduction The idea of organizing the learning sequence of a machine is an old concept that stems from multiple works in reinforcement learning [1, 2] , developmental robotics [3] and supervised learning  ... 
arXiv:2011.08463v3 fatcat:qnxqwfdvj5amvi3tjt733b5vhe


Sutuma Edessa
2015 International Journal Of Biology Education  
academic calendar and the curricula.  ...  The purpose of the study was to assess the range of completed portions in comparison with the omitted ones and investigate the loss or educational wastages with their continuing impacts on the academic  ...  Acknowledgement I deeply express my heart-felt thanks to Professor Tesema Ta'a, a senior professor at Addis Ababa University for his invaluable work to revise and advice in writing up this article.  ... 
doi:10.20876/ijobed.16291 fatcat:ytkfufr4f5gbjhffuwdreeoedu

Open systems metaphor in instructional design

Ali Baykal
2009 Procedia - Social and Behavioral Sciences  
Every kind of stimulation from the external environment of any instructional unit in any form is an input to the system. Money, material, information and people are just a few examples of inputs.  ...  In the information age innovative instructional deign is essential for every country.  ...  They wish to apply the accumulated knowledge by introducing purposeful changes into instructional practices by making them more effective, more efficient for human learning (Biehler, 1986) .  ... 
doi:10.1016/j.sbspro.2009.01.356 fatcat:4ewucnxw3zh4tfiasuc5foijja

Unsupervised Control Through Non-Parametric Discriminative Rewards [article]

David Warde-Farley, Tom Van de Wiele, Tejas Kulkarni, Catalin Ionescu, Steven Hansen, Volodymyr Mnih
2018 arXiv   pre-print
Learning to control an environment without hand-crafted rewards or expert data remains challenging and is at the frontier of reinforcement learning research.  ...  This dual optimization leads to a co-operative game, giving rise to a learned reward function that reflects similarity in controllable aspects of the environment instead of distance in the space of observations  ...  We additionally thank Andriy Mnih and Carlos Florensa for helpful discussions, and Lasse Espeholt, Tom Erez and Dumitru Erhan for invaluable technical assistance.  ... 
arXiv:1811.11359v1 fatcat:ytbyfennwrakfm753umn5ofzwe

Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions [article]

Rui Wang, Joel Lehman, Jeff Clune, Kenneth O. Stanley
2019 arXiv   pre-print
The term open-ended signifies the intriguing potential for algorithms like POET to continue to create novel and increasingly complex capabilities without bound.  ...  Such a process would in effect build its own diverse and expanding curricula, and the solutions to problems at various stages would become stepping stones towards solving even more challenging problems  ...  We also thank Alex Gajewski (from his internship at Uber AI Labs) for help with implementation and useful discussions.  ... 
arXiv:1901.01753v3 fatcat:4pwvjy7pcjb6pfeg5sjahcrvg4

Emergent Tool Use From Multi-Agent Autocurricula [article]

Bowen Baker, Ingmar Kanitscheider, Todor Markov, Yi Wu, Glenn Powell, Bob McGrew, Igor Mordatch
2020 arXiv   pre-print
We find clear evidence of six emergent phases in agent strategy in our environment, each of which creates a new pressure for the opposing team to adapt; for instance, agents learn to build multi-object  ...  We further provide evidence that multi-agent competition may scale better with increasing environment complexity and leads to behavior that centers around far more human-relevant skills than other self-supervised  ...  We also thank Alex Ray for writing parts of our open sourced code.  ... 
arXiv:1909.07528v2 fatcat:u4efcciy7nbl3g5s4kwnillcci

TeachMyAgent: a Benchmark for Automatic Curriculum Learning in Deep RL [article]

Clément Romac, Rémy Portelas, Katja Hofmann, Pierre-Yves Oudeyer
2021 arXiv   pre-print
Training autonomous agents able to generalize to multiple tasks is a key target of Deep Reinforcement Learning (DRL) research.  ...  In parallel to improving DRL algorithms themselves, Automatic Curriculum Learning (ACL) study how teacher algorithms can train DRL agents more efficiently by adapting task selection to their evolving abilities  ...  Jabri, A., Hsu, K., Gupta, A., Eysenbach, B., Levine, S., and Finn, C. Unsupervised curricula for visual meta- reinforcement learning. In Wallach, H.  ... 
arXiv:2103.09815v2 fatcat:qtsp6ghsgrd47fwgznnqdwrtsu

Special Issue on the AMCIS 2001 Workshops: Integrating Enterprise Systems in the University Curriculum

Michael Rosemann, Edward E. Watson
2002 Communications of the Association for Information Systems  
Many interesting attributes are associated with this challenge: students generally have a strong interest to learn the subject and are often biased towards product-focused materials and skills; Enterprise  ...  The ideas in this paper were presented in a special pre-conference workshop at the 2001 annual meeting of the Americas Conference on Information Systems (AMCIS) in Boston.  ...  But also, enterprise systems help students to identify with the real world as they transfer learned concepts and principles from the classroom into real-life business practice and complexity.  ... 
doi:10.17705/1cais.00815 fatcat:6rea4z54h5axlcon7xwudcrnai

Meta-learners' learning dynamics are unlike learners' [article]

Neil C. Rabinowitz
2019 arXiv   pre-print
Meta-learning is a tool that allows us to build sample-efficient learning systems.  ...  In each case, while sample-inefficient DL and RL Learners uncover the task structure in a staggered manner, meta-trained LSTM Meta-Learners uncover almost all task structure concurrently, congruent with  ...  HOW TO ASSESS LEARNING DYNAMICS We thus consider here a Meta-Learner LSTM with fixed weights.  ... 
arXiv:1905.01320v1 fatcat:okuureeiwrblhp7s3nltwojl3i

Why Are Adaptive Learning Organizations Better Placed to Succeed in the Future? Insights from Research on Adaptive Learning Organization by NIIT and John Bersin Academy [chapter]

Kamal Dhuper
2022 China and Globalization  
Learning and Development (L&D) has become more complex and increased pressure on organizations, while Adaptive Learning Organizations (ALOs) have three characteristics of adaptivity that enable a proactive  ...  In the face of such rapid change, executives are piecing together the future landscape of value and the new rules of competitive advantage for their organizations.  ...  This is further being driven by the need for highly personalized, relevant, and efficient learning delivered in ways that resonate with the learner persona.  ... 
doi:10.1007/978-981-16-8603-0_19 fatcat:gnpkveofqvgonjlqyadbmq3sa4
« Previous Showing results 1 — 15 out of 4,692 results