Nonparametric learning rules from bandit experiments: The eyes have it!

Hu, Yingyao; Kayaba, Yutaka; Shum, Matthew

doi:10.1016/j.geb.2013.05.003

Cited by 27 publications

(27 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Regardless of whether a participant perceived a low price ( U t − 1 = 1) or a high price ( U t − 1 = 2), the subjects' beliefs were updated toward certain products with more than a 70% probability and toward uncertain products with an approximate probability of 20%. These results indicate that subjects were more engaged in rational learning, judging, and choice rather than making random choices (choice probabilities between 40% and 60% are characterized as random; Hu, Kayaba, & Shum, ). However, after the information state, the participants entered the tourism image advertising stage S t .…”

Section: Resultsmentioning

confidence: 94%

Consumers' Bayesian learning under Knightian uncertainty: An eye‐tracking analysis

Song

Jiang

2018

J of Consumer Behaviour

View full text Add to dashboard Cite

In recent years, risky decisions and the “gambling paradigm” have been mainstream topics in economics and psychology research. However, the choice of Knightian uncertainty, especially as it relates to the potential cognitive processes of consumers in marketing, has not drawn the attention of scholars. This study introduced Bayesian learning theory to reproduce the consumer learning process and conducted a two‐stage experiment to explore the process of how perceived price and uncertain quality information influence consumer learning and decisions. The results were validated via an eye‐tracking experiment with dynamic (updated) advertising information to observe the learning mechanism of consumers under Knightian uncertainty. The results showed that consumers conduct irregular Bayesian learning based on different perceived price levels in different information state spaces. Belief changes follow the first‐order Markov learning rule, and Bayesian learning is a dual‐system process. Overall, the results are in line with the belief‐integration principle of the Bayesian learning model under Knightian uncertainty.

show abstract

Section: Resultsmentioning

confidence: 94%

Consumers' Bayesian learning under Knightian uncertainty: An eye‐tracking analysis

Song

Jiang

2018

J of Consumer Behaviour

View full text Add to dashboard Cite

show abstract

“…When the sample size becomes larger, the probability of using this remedy should be smaller when all the assumptions hold. This closed-form estimator performs well in empirical studies, such as An, Baye, Hu, Morgan, and Shum (2012), An, Hu, and Shum (2010), Feng and Hu (2013), and Hu, Kayaba, and Shum (2013).…”

Section: Estimationmentioning

confidence: 79%

“…8 Therefore, it doesn't really matter which one of fX t 1 ; X t ; X t+1 g is treated as measurement or instrument for X t . Applications of nonparametric identi…cation of such a hidden Markov model or, in general, the 3-measurement model can be found in Hu, Kayaba, andShum (2013), Feng and, Wilhelm (2013), and Hu and Sasaki (2014), etc.…”

Section: A 3-measurement Modelmentioning

confidence: 99%

“…The key di¢ culty in the estimation of learning models is that beliefs are time-varying and unobserved in the data. Hu, Kayaba, and Shum (2013) use bandit experiments to nonparametrically estimate the learning rule using auxiliary measurements of beliefs. In each period, an economic agent is asked to choose between two slot machines, which have di¤erent winning probabilities.…”

Section: Dynamic Learning Modelsmentioning

confidence: 99%

See 1 more Smart Citation

Microeconomic Models with Latent Variables: Applications of Measurement Error Models in Empirical Industrial Organization and Labor Economics

2015

SSRN Journal

Self Cite

View full text Add to dashboard Cite

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in AbstractThis paper reviews recent developments in nonparametric identi…cation of measurement error models and their applications in applied microeconomics, in particular, in empirical industrial organization and labor economics. Measurement error models describe mappings from a latent distribution to an observed distribution. The identi…cation and estimation of measurement error models focus on how to obtain the latent distribution and the measurement error distribution from the observed distribution. Such a framework may be suitable for many microeconomic models with latent variables, such as models with unobserved heterogeneity or unobserved state variables and panel data models with …xed e¤ects. Recent developments in measurement error models allow very ‡exible speci…cation of the latent distribution and the measurement error distribution. These developments greatly broaden economic applications of measurement error models. This paper provides an accessible introduction of these technical results to empirical researchers so as to expand applications of measurement error models.JEL classi…cation:

show abstract

“…, [25] where ν > 0 is a step-size parameter and b(t) parameters are initialized as bj(0) = logRj. [26] Finally,R is an estimate of irreducible uncertainty.…”

Section: Modeling Learning and Choices -Control Modelsmentioning

confidence: 99%

Uncertainty in learning, choice and visual fixation

Stojić¹,

Orquin²,

Dayan³

et al. 2019

Preprint

View full text Add to dashboard Cite

Uncertainty plays a critical role in reinforcement learning and decision making. However, exactly how it influences behaviour remains unclear. Multi-armed bandit tasks offer an ideal test-bed, since computational tools such as approximate Kalman filters can closely characterize the interplay between trial-by-trial values, uncertainty, learning, and choice. To gain additional insight into learning and choice processes we obtained data from subjects' overt allocation of gaze. The estimated value and estimation uncertainty of options influenced what subjects looked at before choosing; these same quantities also influenced choice, as additionally did fixation itself. A momentary measure of uncertainty in the form of absolute prediction errors determined how long participants looked at the obtained outcomes. These findings affirm the importance of uncertainty in multiple facets of behaviour, and help delineate its effects on decision making.

show abstract

Nonparametric learning rules from bandit experiments: The eyes have it!

Cited by 27 publications

References 57 publications

Consumers' Bayesian learning under Knightian uncertainty: An eye‐tracking analysis

Consumers' Bayesian learning under Knightian uncertainty: An eye‐tracking analysis

Microeconomic Models with Latent Variables: Applications of Measurement Error Models in Empirical Industrial Organization and Labor Economics

Uncertainty in learning, choice and visual fixation

Contact Info

Product

Resources

About