Filters








10,152 Hits in 3.7 sec

Characterizing rational versus exponential learning curves [chapter]

Dale Schuurmans
1995 Lecture Notes in Computer Science  
By addressing a simple non-uniformity in the original analysis, this paper shows how the dichotomy between rational and exponential worst case learning curves can be recovered in the distribution free  ...  Here a learning curve can be de ned to be the expected error of a learner's hypotheses as a function of training sample size.  ...  Comparing uniform versus non-uniform bounds. This demonstrates how a series of exponential learning curves can have a rational upper envelope.  ... 
doi:10.1007/3-540-59119-2_184 fatcat:tnxkpqkdvnbtjbf2kfwtcqcc6e

Characterizing Rational versus Exponential Learning Curves

Dale Schuurmans
1997 Journal of computer and system sciences (Print)  
By addressing a simple non-uniformity in the original analysis this paper shows how the dichotomy between rational and exponential worst case learning curves can be recovered in the distribution-free theory  ...  We consider the standard problem of learning a concept from random examples. Here a learning curve is defined to be the expected error of a learner's hypotheses as a function of training sample size.  ...  Of course, a necessary prerequisite for any practical characterization of empirical learning curves is predicting whether rational versus exponential convergence will take place: obviously one cannot accurately  ... 
doi:10.1006/jcss.1997.1505 fatcat:twy72eqcrvhidcw3t3of733xha

Page 2592 of Mathematical Reviews Vol. , Issue 97D [page]

1997 Mathematical Reviews  
Sanjay Jain (SGP-SING-IS; Singapore) 97d:68192 68T05 Schuurmans, Dale (3-TRNT-C; Toronto, ON) Characterizing rational versus exponential learning curves.  ...  “By addressing a simple non-uniformity in the original anal- ysis, this paper shows how the dichotomy between rational and exponential worst case learning curves can be recovered in the distribution free  ... 

Page 2073 of Mathematical Reviews Vol. , Issue 99c [page]

1991 Mathematical Reviews  
rational versus exponential learning curves.  ...  Summary: “We define distances between geometric curves by the square root of the minimal energy required to transform one curve into the other.  ... 

Page 1901 of Mathematical Reviews Vol. , Issue 97C [page]

1997 Mathematical Reviews  
examples (252-260); Eric Martin and Daniel Osherson, A note on the use of probabilities by mechanical learners (261-271); Dale Schuur- mans, Characterizing rational versus exponential learning curves  ...  Schapire, A decision-theoretic generalization of on-line learning and an application to boosting (23-37); Shai Ben-David, Eyal Kushilevitz and Yishay Mansour, Online learning versus offline learning (38  ... 

Competitive Centipede Games: Zero-End Payoffs and Payoff Inequality Deter Reciprocal Cooperation

Eva Krockow, Briony Pulford, Andrew Colman
2015 Games  
We investigated cooperation in four Centipede games differing in their payoffs at the game's end (positive versus zero) and payoff difference between players (moderate versus high difference).  ...  The learning curves for the different games do not show any discernible trends of either increasing or decreasing cooperation with greater experience in the game.  ...  In a complex research design investigating decision making in Take-it-or-leave-it Centipede games with partially unknown payoff functions, continuous versus discrete moves, and simultaneous versus sequential  ... 
doi:10.3390/g6030262 fatcat:raykxt2dfrhnhogjtkyzlqxeie

Some Thoughts on Reliability of Diagnoses by Human Versus by Machine

Ildikó Ziegler
2020 Global Journal of Engineering Sciences  
It means that with the elapsed time the standard deviation characterizing of the learning process of a given A.I. will decrease during the learning period.  ...  Accidental errors (others than the mentioned in the earlier groups) are characterized by Gaussian distribution [3] [4] [5] [6] .  ...  It will be interesting to see in the long run if the size of the data set used for machine learning -analogously to the afore mentioned laws -will follow an exponential curve as a result of technological  ... 
doi:10.33552/gjes.2020.05.000611 fatcat:mfizkgbldzg2zpqlwoynwhnn6y

Simple Formulae, Deep Learning and Elaborate Modelling for the COVID-19 Pandemic

Athanassios S. Fokas, Nikolaos Dikaios, Sotirios Tsiodras, George A. Kastis
2022 Encyclopedia  
It is emphasized that researchers' forecasting models exhibit, for large t, algebraic behavior, as opposed to the exponential behavior of the classical logistic-type models used usually in epidemics.  ...  Remarkably, a newly introduced mechanistic model also exhibits, for large t, algebraic behavior in contrast to the usual Susceptible-Exposed-Infectious-Removed (SEIR) models, which exhibit exponential  ...  by an algebraic as opposed to an exponential decay.  ... 
doi:10.3390/encyclopedia2020047 fatcat:vfkm4qsiprdmpl3g35sr5d5zsi

Beliefs, Doubts and Learning: Valuing Macroeconomic Risk

Lars Peter Hansen
2007 The American Economic Review  
The − curve was computed assuming that θ b = 24, and --curve was computed assuming that θ b = 6.  ...  The smooth curves in Figure 5 are computed for θ b = 24 and the more volatile curves for θ b = 6.  ... 
doi:10.1257/aer.97.2.1 fatcat:dzld3755mrdghaylsb5tyrgkny

Time scales in motor learning and development

Karl M. Newell, Yeou-Teh Liu, Gottfried Mayer-Kress
2001 Psychological review  
recognition of the importance of rational or theoretical curve fitting versus that based on a mere empirical agenda (Guildford, 1936; Thurstone, 1919) .  ...  nature of the functions of motor learning rather than a narrowing of the learning problem to a theoretical rationale for a single function.  ...  We denote by n the practice time (number of trials) and by x n a variable that characterizes the movement pattern, that is, the associated performance.  ... 
doi:10.1037/0033-295x.108.1.57 pmid:11212633 fatcat:3xpza7o6gndihotkdbi7co6ori

Time scales in motor learning and development

Karl M. Newell, Yeou-Teh Liu, Gottfried Mayer-Kress
2001 Psychological review  
recognition of the importance of rational or theoretical curve fitting versus that based on a mere empirical agenda (Guildford, 1936; Thurstone, 1919) .  ...  nature of the functions of motor learning rather than a narrowing of the learning problem to a theoretical rationale for a single function.  ...  We denote by n the practice time (number of trials) and by x n a variable that characterizes the movement pattern, that is, the associated performance.  ... 
doi:10.1037//0033-295x.108.1.57 fatcat:5qngkzxm3ngh5ndfbzmg7y3mou

Mathematical models and deep learning for predicting the number of individuals reported to be infected with SARS-CoV-2

A. S. Fokas, N. Dikaios, G. A. Kastis
2020 Journal of the Royal Society Interface  
This methodology, which is based on the synergy of explicit mathematical formulae and deep learning networks, yields algorithms whose input is only the existing data in the given country of the accumulative  ...  Figure 4 presents the predictions made by the analytical formulae and the deep learning network versus the actual data for the cumulative number of reported cases due to SARS-CoV-2, as a function of days  ...  Hence, choosing these parameters by requiring that the analytical solution matches the data curve is consistent with the approach of machine learning.  ... 
doi:10.1098/rsif.2020.0494 pmid:32752997 fatcat:vt75pi2b2ndl3on4h7zf7s2dba

Stabilization Theory and Policy: 50 Years after the Phillips Curve

STEPHEN J. TURNOVSKY
2009 Economica  
The learning procedures we have been outlining can be characterized as being "passive", in the sense that the agent learns about the relevant parameters over time as the system evolves and information  ...  Fourth, learning may take different forms, the two most common being least squares learning and Bayesian learning.  ... 
doi:10.1111/j.1468-0335.2009.00807.x fatcat:47piiah5anco5mmuewoki3myke

An Adversarial Interpretation of Information-Theoretic Bounded Rationality [article]

Pedro A. Ortega, Daniel D. Lee
2014 arXiv   pre-print
The adversary can, by paying an exponential penalty, generate costs that diminish the decision maker's payoffs.  ...  IT bounded rationality addresses this question by defining a new objective function that trades off utilities versus information costs.  ...  Given a function f (x), the convex conjugate f (s) corresponds to the intercept of a tangent line to the curve with slope s. 4.  ... 
arXiv:1404.5668v1 fatcat:tqhmsfibb5edzcsh2tqtijkvlq

An Adversarial Interpretation of Information-Theoretic Bounded Rationality

Pedro Ortega, Daniel Lee
2014 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
The adversary can, by paying an exponential penalty, generate costs that diminish the decision maker's payoffs.  ...  IT bounded rationality addresses this question by defining a new objective function that trades off utilities versus information costs.  ...  Given a function f (x), the convex conjugate f (s) corresponds to the intercept of a tangent line to the curve with slope s. 4.  ... 
doi:10.1609/aaai.v28i1.9071 fatcat:yn3pw2ko5vddtowxq7rnh6psp4
« Previous Showing results 1 — 15 out of 10,152 results