The Experts below are selected from a list of 144828 Experts worldwide ranked by ideXlab platform
Matthew D Taylor - One of the best experts on this subject based on the ideXlab platform.
-
Curriculum Design for machine learners in sequential decision tasks
IEEE Transactions on Emerging Topics in Computational Intelligence, 2018Co-Authors: Bei Peng, James Macglashan, Robert Loftin, Michael L Littman, David L Roberts, Matthew D TaylorAbstract:Existing work in machine learning has shown that algorithms can benefit from the use of curricula—learning first on simple examples before moving to more difficult problems. This work studies the Curriculum-Design problem in the context of sequential decision tasks, analyzing how different curricula affect learning in a Sokoban-like domain, and presenting the results of a user study that explores whether nonexperts generate effective curricula. Our results show that 1) the way in which evaluative feedback is given to the agent as it learns individual tasks does not affect the relative quality of different curricula, 2) nonexpert users can successfully Design curricula that result in better overall performance than having the agent learn from scratch, and 3) nonexpert users can discover and follow salient principles when selecting tasks in a Curriculum. We also demonstrate that our Curriculum-learning algorithm can be improved by incorporating the principles people use when Designing curricula. This work gives us insights into the development of new machine-learning algorithms and interfaces that can better accommodate machine- or human-created curricula.
-
Curriculum Design for machine learners in sequential decision tasks
Adaptive Agents and Multi-Agents Systems, 2017Co-Authors: Bei Peng, James Macglashan, Robert Loftin, Michael L Littman, David L Roberts, Matthew D TaylorAbstract:Existing machine-learning work has shown that algorithms can benefit from curricula---learning first on simple examples before moving to more difficult examples. While most existing work on Curriculum learning focuses on developing automatic methods to iteratively select training examples with increasing difficulty tailored to the current ability of the learner, relatively little attention has been paid to the ways in which humans Design curricula. We argue that a better understanding of the human-Designed curricula could give us insights into the development of new machine-learning algorithms and interfaces that can better accommodate machine- or human-created curricula. Our work addresses this emerging and vital area empirically, taking an important step to characterize the nature of human-Designed curricula relative to the space of possible curricula and the performance benefits that may (or may not) occur.
Michael Kyte - One of the best experts on this subject based on the ideXlab platform.
-
a model for collaborative Curriculum Design in transportation engineer ing education
2013 ASEE Annual Conference & Exposition, 2013Co-Authors: Kristen Sanford L Bernhardt, David S Hurwitz, Rhonda Young, Rod E Turochy, Shane Brown, Joshua Swake, Andrea R Bill, Kevin Heaslip, Michael KyteAbstract:The National Transportation Curriculum Project (NTCP) has been underway for four years as an ad-hoc, collaborative effort to effect changes in transportation engineering education. Specifically, the NTCP had developed a set of learning outcomes and associated knowledge tables for the introductory transportation engineering course that is taught in most civil engineering programs, and most recently the project led a workshop, supported by the National Science Foundation, in which approximately 60 participants developed learning and assessment activities to support these learning outcomes. The inter-generational, geographically and institutionally diverse group of faculty members that form the core project group provide a model for cross-institutional collaborative Curriculum Design.
Bei Peng - One of the best experts on this subject based on the ideXlab platform.
-
Curriculum Design for machine learners in sequential decision tasks
IEEE Transactions on Emerging Topics in Computational Intelligence, 2018Co-Authors: Bei Peng, James Macglashan, Robert Loftin, Michael L Littman, David L Roberts, Matthew D TaylorAbstract:Existing work in machine learning has shown that algorithms can benefit from the use of curricula—learning first on simple examples before moving to more difficult problems. This work studies the Curriculum-Design problem in the context of sequential decision tasks, analyzing how different curricula affect learning in a Sokoban-like domain, and presenting the results of a user study that explores whether nonexperts generate effective curricula. Our results show that 1) the way in which evaluative feedback is given to the agent as it learns individual tasks does not affect the relative quality of different curricula, 2) nonexpert users can successfully Design curricula that result in better overall performance than having the agent learn from scratch, and 3) nonexpert users can discover and follow salient principles when selecting tasks in a Curriculum. We also demonstrate that our Curriculum-learning algorithm can be improved by incorporating the principles people use when Designing curricula. This work gives us insights into the development of new machine-learning algorithms and interfaces that can better accommodate machine- or human-created curricula.
-
Curriculum Design for machine learners in sequential decision tasks
Adaptive Agents and Multi-Agents Systems, 2017Co-Authors: Bei Peng, James Macglashan, Robert Loftin, Michael L Littman, David L Roberts, Matthew D TaylorAbstract:Existing machine-learning work has shown that algorithms can benefit from curricula---learning first on simple examples before moving to more difficult examples. While most existing work on Curriculum learning focuses on developing automatic methods to iteratively select training examples with increasing difficulty tailored to the current ability of the learner, relatively little attention has been paid to the ways in which humans Design curricula. We argue that a better understanding of the human-Designed curricula could give us insights into the development of new machine-learning algorithms and interfaces that can better accommodate machine- or human-created curricula. Our work addresses this emerging and vital area empirically, taking an important step to characterize the nature of human-Designed curricula relative to the space of possible curricula and the performance benefits that may (or may not) occur.
Barrie Margetts - One of the best experts on this subject based on the ideXlab platform.
-
Curriculum Design for professional development in public health nutrition in britain
Public Health Nutrition, 1998Co-Authors: Jacqueline Landman, Judith Buttriss, Barrie MargettsAbstract:OBJECTIVES To describe how the Nutrition Society developed public health nutrition as a profession between 1992 and 1997, and to analyse the influences propelling on this professionalization. Design Qualitative case study. SETTING Britain. RESULTS The Nutrition Society of Britain consulted with various stakeholders (such as dietitians, researchers, professionals and practitioners and educators from the UK, and latterly from mainland Europe) to build a consensus about the definition, roles and functions of public health nutritionists and the need for, and scope of, this new profession. Building on this consensus, the Society developed a Curriculum in line with British national nutrition policy. Analysis shows that the Design and philosophy of the Curriculum is explicitly international and European in orientation, in keeping with the tradition of the discipline and the Society. The Curriculum is Designed in terms of specialist competencies in public health nutrition, defining competency so that registered public health nutritionists are advanced practitioners or leaders: this is in keeping with contemporary trends in professional education generally and as expressed by the UNU/IUNS and at Bellagio, in nutrition in particular. CONCLUSIONS Despite a unique relationship with British state and policy, this case of professionalization contributes to contemporary international inter- and intraprofessional debates about the nature of public health nutrition and is consistent with professional educational theory.
Michael L Littman - One of the best experts on this subject based on the ideXlab platform.
-
Curriculum Design for machine learners in sequential decision tasks
IEEE Transactions on Emerging Topics in Computational Intelligence, 2018Co-Authors: Bei Peng, James Macglashan, Robert Loftin, Michael L Littman, David L Roberts, Matthew D TaylorAbstract:Existing work in machine learning has shown that algorithms can benefit from the use of curricula—learning first on simple examples before moving to more difficult problems. This work studies the Curriculum-Design problem in the context of sequential decision tasks, analyzing how different curricula affect learning in a Sokoban-like domain, and presenting the results of a user study that explores whether nonexperts generate effective curricula. Our results show that 1) the way in which evaluative feedback is given to the agent as it learns individual tasks does not affect the relative quality of different curricula, 2) nonexpert users can successfully Design curricula that result in better overall performance than having the agent learn from scratch, and 3) nonexpert users can discover and follow salient principles when selecting tasks in a Curriculum. We also demonstrate that our Curriculum-learning algorithm can be improved by incorporating the principles people use when Designing curricula. This work gives us insights into the development of new machine-learning algorithms and interfaces that can better accommodate machine- or human-created curricula.
-
Curriculum Design for machine learners in sequential decision tasks
Adaptive Agents and Multi-Agents Systems, 2017Co-Authors: Bei Peng, James Macglashan, Robert Loftin, Michael L Littman, David L Roberts, Matthew D TaylorAbstract:Existing machine-learning work has shown that algorithms can benefit from curricula---learning first on simple examples before moving to more difficult examples. While most existing work on Curriculum learning focuses on developing automatic methods to iteratively select training examples with increasing difficulty tailored to the current ability of the learner, relatively little attention has been paid to the ways in which humans Design curricula. We argue that a better understanding of the human-Designed curricula could give us insights into the development of new machine-learning algorithms and interfaces that can better accommodate machine- or human-created curricula. Our work addresses this emerging and vital area empirically, taking an important step to characterize the nature of human-Designed curricula relative to the space of possible curricula and the performance benefits that may (or may not) occur.
