scholarly journals Validating model-based Bayesian integration using prior–cost metamers

2021 ◽  
Vol 118 (25) ◽  
pp. e2021531118
Author(s):  
Hansem Sohn ◽  
Mehrdad Jazayeri
Keyword(s):  
Bayesian Learning ◽  
Learning Strategy ◽  
Alternative Hypothesis ◽  
Internal Models ◽  
Cost Functions ◽  
Bayesian Decision ◽  
Visuomotor Rotation ◽  
Model Based ◽  
Sensory Experiences ◽  
Timing Task

There are two competing views on how humans make decisions under uncertainty. Bayesian decision theory posits that humans optimize their behavior by establishing and integrating internal models of past sensory experiences (priors) and decision outcomes (cost functions). An alternative hypothesis posits that decisions are optimized through trial and error without explicit internal models for priors and cost functions. To distinguish between these possibilities, we introduce a paradigm that probes the sensitivity of humans to transitions between prior–cost pairs that demand the same optimal policy (metamers) but distinct internal models. We demonstrate the utility of our approach in two experiments that were classically explained by Bayesian theory. Our approach validates the Bayesian learning strategy in an interval timing task but not in a visuomotor rotation task. More generally, our work provides a domain-general approach for testing the circumstances under which humans explicitly implement model-based Bayesian computations.

scholarly journals Metamers of Bayesian computation

2020 ◽  
Author(s):  
Hansem Sohn ◽  
Mehrdad Jazayeri
Keyword(s):  
Internal Models ◽  
Cost Functions ◽  
Trial And Error ◽  
Bayesian Decision ◽  
Visuomotor Rotation ◽  
Model Based ◽  
Model Free ◽  
Sensory Experiences ◽  
Timing Task ◽  
Decision Outcomes

AbstractThere are two sharply debated views on how humans make decisions under uncertainty. Bayesian decision theory posits that humans optimize their behavior by establishing and integrating internal models of past sensory experiences (priors) and decision outcomes (cost functions). An alternative model-free hypothesis posits that decisions are optimized through trial and error without explicit internal models for priors and cost functions. To distinguish between these possibilities, we introduce a novel paradigm that probes sensitivity of humans to transitions between prior-cost pairs that demand the same optimal policy (metamers) but distinct internal models. We demonstrate the utility of our approach in two experiments that were classically explained by model-based Bayesian theory. Our approach validates the model-based strategy in an interval timing task but not in a visuomotor rotation task. More generally, our work provides a domain-general approach for testing the circumstances under which humans implement model-based Bayesian computations.

Representation, abstraction, and simple-minded sophisticates

2020 ◽  
Vol 43 ◽  
Author(s):  
Peter Dayan
Keyword(s):  
Decision Theory ◽  
Predictive Coding ◽  
Bayesian Decision Theory ◽  
Bayesian Decision ◽  
Model Based ◽  
Model Free

Abstract Bayesian decision theory provides a simple formal elucidation of some of the ways that representation and representational abstraction are involved with, and exploit, both prediction and its rather distant cousin, predictive coding. Both model-free and model-based methods are involved.

scholarly journals Severe Sexual Abuse Reduces Frontoparietal Network Activity during Model-Based Reinforcement Learning Updates

2019 ◽  
Author(s):  
Allison Letkiewicz ◽  
Amy L. Cochran ◽  
Josh M. Cisler
Keyword(s):  
Sexual Abuse ◽  
Reinforcement Learning ◽  
Learning Styles ◽  
Computational Models ◽  
Learning Strategy ◽  
Network Activity ◽  
Model Based ◽  
Model Free ◽  
Frontoparietal Network ◽  
Assaultive Trauma

Trauma and trauma-related disorders are characterized by altered learning styles. Two learning processes that have been delineated using computational modeling are model-free and model-based reinforcement learning (RL), characterized by trial and error and goal-driven, rule-based learning, respectively. Prior research suggests that model-free RL is disrupted among individuals with a history of assaultive trauma and may contribute to altered fear responding. Currently, it is unclear whether model-based RL, which involves building abstract and nuanced representations of stimulus-outcome relationships to prospectively predict action-related outcomes, is also impaired among individuals who have experienced trauma. The present study sought to test the hypothesis of impaired model-based RL among adolescent females exposed to assaultive trauma. Participants (n=60) completed a three-arm bandit RL task during fMRI acquisition. Two computational models compared the degree to which each participant’s task behavior fit the use of a model-free versus model-based RL strategy. Overall, a greater portion of participants’ behavior was better captured by the model-based than model-free RL model. Although assaultive trauma did not predict learning strategy use, greater sexual abuse severity predicted less use of model-based compared to model-free RL. Additionally, severe sexual abuse predicted less left frontoparietal network encoding of model-based RL updates, which was not accounted for by PTSD. Given the significant impact that sexual trauma has on mental health and other aspects of functioning, it is plausible that altered model-based RL is an important route through which clinical impairment emerges.

scholarly journals Penerapan Strategi Pembelajaran Circuit Learning: Suatu Upaya dalam Meningkatkan Hasil Belajar Siswa

2019 ◽  
Vol 9 (2) ◽  
pp. 285
Author(s):  
Syahrial Syahrial
Keyword(s):  
Learning Outcomes ◽  
Learning Strategy ◽  
Alternative Hypothesis ◽  
Mathematics Learning ◽  
Control Group ◽  
Test Analysis ◽  
The Real ◽  
Randomized Control ◽  
Affective Domains ◽  
Better Than

This research is motivated by the low student mathematics learning outcomes. The influencing factors are inactive students and lack of communication between students and students. This study aims to determine the effect of the application of the circuit learning strategy to students' learning outcomes in the cognitive and effective domains. This type of research is pre-experimental and the research design used is randomized control group only design. Based on the final test of learning outcomes obtained an average of mathematics learning outcomes in the experimental class 79.3 and the average mathematics learning outcomes of the control class 70. The results of the t-test analysis obtained tcount = 3.89 and ttable = 1.667 at the real level of 0.05. It is concluded that tcount> ttable accepts an alternative hypothesis (H1) that is the mathematics learning outcomes of the experimental class students is better than the control class. Analysis of the data in the affective domain obtained zcount = 3.83 and ztable = 1.64 at the real level of 0.05 thus zcount> ztable, in other words Hi is accepted meaning that student learning activities in the experimental class are better than the control class. Based on data analysis in the cognitive and affective domains it can be concluded that there is an influence of the application of the circuit learning strategy to student mathematics learning outcomes.

scholarly journals Dealing with Drift Uncertainty: A Bayesian Learning Approach

Risks ◽  
2019 ◽  
Vol 7 (1) ◽  
pp. 5 ◽  
Author(s):  
Carmine De Franco ◽  
Johann Nicolle ◽  
Huyên Pham
Keyword(s):  
Bayesian Learning ◽  
Learning Strategy ◽  
Value Added ◽  
Learning Approach ◽  
Practical Procedure ◽  
Asset Classes ◽  
Two Parameters ◽  
Markowitz Problem ◽  
Drift Uncertainty ◽  
Smart Beta

One of the main challenges investors have to face is model uncertainty. Typically, the dynamic of the assets is modeled using two parameters: the drift vector and the covariance matrix, which are both uncertain. Since the variance/covariance parameter is assumed to be estimated with a certain level of confidence, we focus on drift uncertainty in this paper. Building on filtering techniques and learning methods, we use a Bayesian learning approach to solve the Markowitz problem and provide a simple and practical procedure to implement optimal strategy. To illustrate the value added of using the optimal Bayesian learning strategy, we compare it with an optimal nonlearning strategy that keeps the drift constant at all times. In order to emphasize the prevalence of the Bayesian learning strategy above the nonlearning one in different situations, we experiment three different investment universes: indices of various asset classes, currencies and smart beta strategies.

THE EFFECT OF PROBLEM BASED LEARNING ON STUDENTS’ RECOUNT TEXT WRITING ACHIEVEMENT

2015 ◽  
Vol 4 (1) ◽  
Author(s):  
Dyan Yosephin Hutagalung ◽  
Lince Sihombing ◽  
Elia Masa Gintings
Keyword(s):  
Learning Strategy ◽  
Alternative Hypothesis ◽  
Problem Based Learning ◽  
Control Group ◽  
Writing Achievement ◽  
Test Form ◽  
The Mean ◽  
Written Test ◽  
Experimental Group ◽  
Text Writing

This research deals with the effect of problem based learning on students’ recount text writing achievement. The population of this research was the students of  VIII SMP of SMP Negeri 18 Medan in Academic Year 2015 / 2016. This research was conducted with two randomized groups namely experimental group and control group. The experimental group was taught by using problem based learning strategy and the control group was taught without using problem based learning. The researcher collected the data by giving essay test in the written test form. In order to know the reliability of the test, the researcher used inter-rater formula. The result of the reliability was 0, 93. The data was analyzed by using t-test formula. The analysis showed that the mean scores of the students in the experimental group was significantly higher than the mean scores of the students in the control group at the level of significant α= 0.05 with the degree of freedom (df) 70 with tobserved value 7,327 > ttable value 1,994. Therefore, null hypothesis (Ho) was rejected and alternative hypothesis (Ha) was accepted. The finding indicated that problem based learning  significantly affected the students’ achievement in writing recount text. Key words: Problem Based Learning, Recount Text, Writing Achievement

Pengaruh Pengunaan Model Jigsaw Terhadap Kemampuan Mengidentifikasi Teks Ceramah Siswa Kelas XI SMK Negeri 3 Padang

Horizon ◽  
2021 ◽  
Vol 1 (3) ◽  
pp. 464-471
Author(s):  
Amelia Fitria Sari ◽  
Yulia Sri Hartati ◽  
Asri Wahyuni Sari
Keyword(s):  
Data Analysis ◽  
Quantitative Research ◽  
Alternative Hypothesis ◽  
Control Design ◽  
Learning Model ◽  
Average Score ◽  
Test Results ◽  
Model Based ◽  
Model Class

This research is motivated by the problem that some students do not understand the material about identifying lecture texts. The purpose of this study was to describe the ability to identify lecture texts for class XI students at SMK N 3 Padang without and using the Jigsaw learning model, and to describe the effect of the Jigsaw learning model on the ability to identify lecture texts for class XI students at SMK N 3 Padang. This type of research is quantitative research. The design of this study was a posttest only control design. The population in this study were class XI students of SMK N 3 Padang who were registered in 2019/2020. The research sample consisted of 30 control classes and 30 sample classes. The data of this study is the score of the test results of the ability to identify lecture texts without and using the Jigsaw learning model. Based on the results of the study, it was concluded as follows: First, the level of Ability to Identify Lecture Texts without using the Jigsaw model, class XI students of SMK N 3 Padang obtained an average score of 71.30 with a classification of 66-75%, which is more than adequate (LdC). Second, the level of ability to identify lecture texts after using the Jigsaw model, class XI students of SMK N 3 Padang obtained an average score of 79.68 with a classification of 76-85%, which is good (B). Third, from the results of data analysis that has been carried out that the use of the Jigsaw model has a significant effect on increasing the ability to identify lecture texts, it can be seen that the alternative hypothesis (H1) is accepted at a significant level of 95% and dk = n-2 because tcount > ttable ( 5.17 > 1.67).

Sign in / Sign up

Export Citation Format

Share Document