Filters








106,165 Hits in 5.8 sec

Deciding What to Learn: A Rate-Distortion Approach [article]

Dilip Arumugam, Benjamin Van Roy
2021 arXiv   pre-print
We establish a general bound on expected discounted regret for an agent that decides what to learn in this manner along with computational experiments that illustrate the expressiveness of designer preferences  ...  In this work, leveraging rate-distortion theory, we automate this process such that the designer need only express their preferences via a single hyperparameter and the agent is endowed with the ability  ...  We introduce in this paper what is possibly the first principled approach to address a fundamental question: how should an agent decide what to learn?  ... 
arXiv:2101.06197v3 fatcat:4jmrhe2qxnc4bhlzltw4r77edy

Deciding What to Model: Value-Equivalent Sampling for Reinforcement Learning [article]

Dilip Arumugam, Benjamin Van Roy
2022 arXiv   pre-print
To address this problem, we introduce an algorithm that, using rate-distortion theory, iteratively computes an approximately-value-equivalent, lossy compression of the environment which an agent may feasibly  ...  is simple enough to learn while only incurring bounded sub-optimality.  ...  consideration for information-theoretic approaches to guiding representation learning [Abel et al., 2019 , Shafieepoorfard et al., 2016] as a proxy to engaging with a rate-distortion trade-off.  ... 
arXiv:2206.02072v1 fatcat:u5oinjr55bap3grnbuqkoyz3am

Automating Program Speedup by Deciding What to Cache

Jack Mostow, Donald Cohen
1985 International Joint Conference on Artificial Intelligence  
A common program optimization strategy is to eliminate recomputation by caching and reusing results.  ...  We analyze the problems involved in automating this strategy: deciding which computations are safe to cache, transforming the rest of the program to make them safe, choosing the most cost-effective ones  ...  Acknowledgements We would like to thank Bill Swartout for asking some good questions and Bob Balzer for suggesting this problem in the first place.  ... 
dblp:conf/ijcai/MostowC85 fatcat:k6lznrld55dyfkeynah7uo2izu

Obstetricians and pregnant women from Formiga's town, in Minas Gerais State, Brazil. Deciding what kind of labor to be held

Heslley Machado Silva, Angélica Rodrigues da Costa, Radassa de Avelar Nogueira Herculano
2013 Open Journal of Obstetrics and Gynecology  
lead them to decide what kind of delivery should be taken.  ...  The decision of what kind of labor that should be held is intermittently generating a great debate in Brazil.  ...  Further that, there would be a devaluation of learning assistance to the normal delivery [13] .  ... 
doi:10.4236/ojog.2013.31018 fatcat:i5lehcwogjgennpuxm35w2joa4

Digitally Nudged Learning

Jason Richard Byrne, Takehiko Ito, Mariko Furuyabu
2022 International Journal of Emerging Technologies in Learning (iJET)  
The objective was to digitally nudge students towards a gamified online learning tool, thereby improving quiz test performance through a fun motivating language learning game.  ...  A mixed methods approach: the primary quasi-experimental methodology was regression discontinuity design, with a follow up survey.  ...  Acknowledgment We would like to thank Fujishiro, Jinbou and Masui sensei for their kind co-operation.  ... 
doi:10.3991/ijet.v17i12.30567 doaj:e9cd95e5a879421284606484c9c06a7d fatcat:5dt5ynahqjaalamkchc64w2jmy

Multimodal Learning Analytics to Inform Learning Design: Lessons Learned from Computing Education

Katerina Mangaroska, Kshitij Sharma, Dragan Gašević, Michalis Giannakos
2020 Journal of Learning Analytics  
of educators from a post-evaluation design-aware process to a permanent monitoring process of adaptation.  ...  Programming is a complex learning activity that involves coordination of cognitive processes and affective states.  ...  Funding This work was supported by the Research Council of Norway under the project FUTURE LEARNING (255129/H20).  ... 
doi:10.18608/jla.2020.73.7 fatcat:7lmosedkqbbcboyr565o4feh5u

Distinguishing prototype-based and exemplar-based processes in dot-pattern category learning

J. David Smith, John Paul Minda
2002 Journal of Experimental Psychology. Learning, Memory and Cognition  
In Experiments 1A and 1B, participants provided similarity ratings of dot-distortion pairs that were distortions of the same originating prototype.  ...  The results show that comparisons to training exemplars surrounding the prototype create flat typicality gradients within a category and small prototypeenhancement effects, whereas comparisons to a prototype  ...  Your job is to look at each pair and decide how different they are. If there is no difference between them, give them a rating of '1.'  ... 
doi:10.1037/0278-7393.28.4.800 pmid:12109770 fatcat:6ntqwmlskvbuzowhrrb3oeprpm

Learning What to Want: Context-Sensitive Preference Learning

Nisheeth Srivastava, Paul Schrater, Pablo Brañas-Garza
2015 PLoS ONE  
Our method infers preferences that are rational in a psychological sense, where agent choices result from Bayesian inference of what to do from observable inputs.  ...  We have developed a method for learning relative preferences from histories of choices made, without requiring an intermediate utility computation.  ...  A brief note about terminology: throughout this paper, we use the term desirability to denote a learned sense of what to do.  ... 
doi:10.1371/journal.pone.0141129 pmid:26496645 pmcid:PMC4619741 fatcat:upaicivozzeprgldc2u6huoavi

Mastery Learning [chapter]

Thomas R. Guskey
2015 International Encyclopedia of the Social & Behavioral Sciences  
It discusses the improvements in student learning that typically result from the use of mastery learning and how this strategy provides practical solutions to a variety of persistent instructional problems  ...  This article describes how mastery learning originated and the essential elements involved in its implementation.  ...  First, they need to decide what concepts or skills are most important for students to learn and most central to students' understanding.  ... 
doi:10.1016/b978-0-08-097086-8.26039-x fatcat:qncwrvu4nnfilnxzpoys5d55eu

Mastery Learning [chapter]

T.R. Guskey
2001 International Encyclopedia of the Social & Behavioral Sciences  
It discusses the improvements in student learning that typically result from the use of mastery learning and how this strategy provides practical solutions to a variety of persistent instructional problems  ...  This article describes how mastery learning originated and the essential elements involved in its implementation.  ...  First, they need to decide what concepts or skills are most important for students to learn and most central to students' understanding.  ... 
doi:10.1016/b0-08-043076-7/02429-3 fatcat:kzz6ptitefdujmxur6lfvvl4eu

On learning mathematics

Jerome S. Bruner
1960 Mathematics Teacher  
What places do discovery, intuition, and translation have in the learning of mathematics?  ...  Are our present activity and efforts to improve mathematics teaching the beginning of an educational renaissance?  ...  In content, positive knowledge is increasing at a rate that, from the point of view of what por- tion of it one man can know in his life- time, is, to some, alarming.  ... 
doi:10.5951/mt.53.8.0610 fatcat:lqc62l3swvel5jxxq2yiasjrym

China [chapter]

Justin Yifu Lin, Jun Zhang
2019 How Nations Learn  
Late latecomers following the NSE approach and accumulating physical and human capital through learning are most likely to achieve rapid development and to upgrade from imitation to innovation.  ...  chain, paving the way for institutional reform and turning a relatively closed economy into a global manufacturing powerhouse.  ...  What can be Learned from China's Catch-up Learning Experience?  ... 
doi:10.1093/oso/9780198841760.003.0008 fatcat:viit73qbcjgmva5rx5yigjl2vu

Learning-Theoretic Methods in Vector Quantization [chapter]

T. Linder
2002 Principles of Nonparametric Learning  
The goal is to "learn" the optimal quantizer from the data, i.e., to produce empirically designed quantizers with performance approaching (as n gets large) the performance of a quantizer optimal for X.  ...  The strength of the Lagrangian approach is evident here; no such general result is known for variable-rate quantizers that minimize the distortion for a given rate constraint.  ... 
doi:10.1007/978-3-7091-2568-7_4 fatcat:q3kdytcvlfhzxmcpt2vqj5ckce

What can analytics contribute to accessibility in e-learning systems and to disabled students' learning?

Martyn Cooper, Rebecca Ferguson, Annika Wolff
2016 Proceedings of the Sixth International Conference on Learning Analytics & Knowledge - LAK '16  
A comparative analysis of completion rates of disabled and non-disabled students in a large five-year dataset is presented and a wide variation in comparative retention rates is characterized.  ...  This paper is intended to stimulate a wider interest in the potential benefits of learning analytics for institutions as they try to assure the accessibility of their e-learning and provision of support  ...  This approach points to cases where significant accessibility issues may exist; it says nothing about what those accessibility barriers might be.  ... 
doi:10.1145/2883851.2883946 dblp:conf/lak/CooperFW16 fatcat:sa76nco7ffblzfb4owgz7zdenm

Perceptual Quality-preserving Black-Box Attack against Deep Learning Image Classifiers [article]

Diego Gragnaniello, Francesco Marra, Giovanni Poggi, Luisa Verdoliva
2020 arXiv   pre-print
In this work, we propose to perform the black-box attack along a low-distortion path, so as to improve both the attack efficiency and the perceptual quality of the adversarial image.  ...  Numerical experiments on real-world systems prove the effectiveness of the proposed approach, both in benchmark classification tasks and in key applications in biometrics and forensics.  ...  Note that, contrary to what happens with L p distortion laws, perturbations cause a smaller or larger increase in distortion depending on the local context.  ... 
arXiv:1902.07776v3 fatcat:65tu36i6trbitnk5fbwxqsz3v4
« Previous Showing results 1 — 15 out of 106,165 results