Effect of human prior knowledge on game success and comparison with reinforcement learning

Hasanoğlu, Mert.
This study aims to find out the effect of prior knowledge on the success of humans in a non-rewarding game environment, and then to compare human performance with a reinforcement learning method in an effort to observe to what extent this method can be brought closer to human behavior and performance with the data obtained. For this purpose, different versions of a simple 2D game were used, and data were collected from 32 participants. At the end of the experiment, it is concluded that prior knowledge, such as the meaning of objects and colors, have an impact on human performance. It was observed that the reinforcement learning agent failed to finish the same game. Various attempts have been made to improve performance and to achieve human-like behavior. In one of these, mini-games were prepared to intro- duce prior knowledge of the objects and the interaction with them. In another trial, a model is created with the game data collected from participants, and the agent is trained using this model as an exploration strategy. Only when the human data is used as an exploration strategy, the agent succeeded in finishing the game. Although the performance of the reinforcement algorithm is increased, human-like behavior is not observed. The conclusion is that it is more meaningful to consider prior knowledge within the context of exploration strategy, and having prior knowledge is not enough in achieving human-like behavior.


Effects of Content Balancing and Item Selection Method on Ability Estimation in Computerized Adaptive Tests
Sahin, Alper; ÖZBAŞI, DURMUŞ (2017-01-01)
Purpose: This study aims to reveal effects of content balancing and item selection method on ability estimation in computerized adaptive tests by comparing Fisher's maximum information (FMI) and likelihood weighted information (LWI) methods. Research Methods: Four groups of examinees (250, 500, 750, 1000) and a bank of 500 items with 10 different content domains were generated through Monte Carlo simulations. Examinee ability was estimated by fixing all settings except for the item selection methods mention...
Linking EFL learning environment characteristics to persistence in EFL learning: a mixed-design study
Mutlu, Gülçin; Güneri, Oya; Department of Educational Sciences (2017)
This study aims to investigate learning environment characteristics that may relate to persistence in EFL learning as a student affective outcome through qualitative and quantitative facets. An EFL Learning Environment Questionnaire and a Persistence Scale in EFL learning were used to investigate learning environment characteristics and persistence. In addition, qualitative interviews were carried out to describe these characteristics and persistence in EFL learning. The sample for this mixed-design study i...
Effect of Visual Context Information for Super Resolution Problems
Akar, Gözde; Aykut, Ekin; Cengiz, Baran; Bocek, Kadircan (2019-04-26)
In this study, the effect of visual context information to the performance of learning-based techniques for the super resolution problem is analyzed. Beside the interpretation of the experimental results in detail, its theoretical reasoning is also achieved in the paper. For the experiments, two different visual datasets composed of natural and remote sensing scenes are utilized. From the experimental results, we observe that keeping visual context information in the course of parameter learning for convolu...
How Signaling Principle Affects Learning: An Eye Tracking Study
Dönmez, Mehmet; Doğan, Sibel; Baran, Evrim (2018-08-01)
Studies examining the effects of signaling principle on learning mostly used data sources such as interviews, achievement tests and think aloud protocols. There is, however, limited research on the use of quantitative measures to promote findings from interviews and think aloud procedures. This study aimed to investigate the effects of signaling principle on learning by supporting the results with the eye movement data. Participants included 34 Information and Communication Technology (ICT) pre-service teac...
Effects of Combined Peer-Teacher Feedback on Second Language Writing Development
Demirel, Elif; Enginarlar, Hüsnü (2016-10-01)
This study attempts to test the effectiveness of a feedback model which combines teacher and peer feedback systematically on improving students' writing ability in the context of a multiple draft writing course. 57 Turkish EFL students participated in the study which lasted for a 15-week semester. The experimental and control groups were provided different feedback treatments and then statistically compared in terms of the revisions they made in their essay drafts and their writing improvement. A total of 1...
Citation Formats
M. Hasanoğlu, “Effect of human prior knowledge on game success and comparison with reinforcement learning,” Thesis (M.S.) -- Graduate School of Informatics. Cognitive Sciences., Middle East Technical University, 2019.