Finite Memory · Finite Memory and Dynamic Decision Problems
„Хоризонт 2020“ — Действия „Мария Склодовска-Кюри“
- Период
- 2022-06-01 → 2025-05-31
- Финансиране от ЕС
- 237 944 €
- Участници
- 2
- Схема
- MSCA-IF
Линиите свързват координатора с партньорите.
Накратко на български
Математическите модели на избора изследват защо хора с различни предпочитания могат да вземат едни и същи решения. Това помага за по-доброто разбиране на потребителското поведение и начина, по който информацията влияе върху изборите ни.
Кратко обяснение, генерирано от езиков модел по текста на CORDIS. Оригиналът е по-долу.
Резултати накратко
Finite Memory and Dynamic Decision Problems
1. Problem/Issue: Random choice models are often underidentified, meaning that the observed choice behavior can render diverse representations within the model. This makes it challenging to interpret heterogeneity in the population. Prominent random choice models like the random utility model often suffer from underidentification the same observed choice data can be represented in many different ways within the model. This underidentification makes it difficult to uniquely identify and interpret the sources of heterogeneity driving the observed random choice behavior in the population. 2. Importance for society: Understanding heterogeneity in choice behavior is crucial for various economic applications, such as analyzing consumer preferences, risk attitudes, social preferences, and the effects of policies or information on decision-making. It allows for deriving more accurate positive and normative insights about how people's preferences, bounded rationality, cognitive biases etc. impact their choices. This can inform better product design, marketing strategies, policy interventions aimed at nudging behavior, and more. 3. Overall Objectives: - Propose "self-progressive" choice models that guarantee each aggregate choice behavior has a unique orderly representation. - Establish a lattice-theoretic equivalence that provides a recipe to make any choice model self-progressive. - Characterize the minimal self-progressive extension of rational choice - requires choices to satisfy intuitive axioms like respecting removal/addition of inferior/superior options. - Derive conditions under which the primitive ordering over alternatives (reflecting preferences, cognition etc.) can be uniquely identified from choice data. - Explore universally self-progressive models that yield unique representations for any primitive ordering, enabling richer behavioral specifications. - Illustrate the framework using examples capturing phenomena like choice overload, similarity-based choice, multi-criteria decision making etc.
Текст от CORDIS, на английски · Данни: CORDIS, © Европейски съюз
Цел на проекта
Memory limitations lie at the core of how agents store past information and make decisions. For example, consider a buyer choosing between two alternative brands of a good. In thinking about a brand, she may recall her past experiences, product reviews, or expert’s ratings, which induce a belief about the brand’s quality. In this process, the buyer may divide the quality scale into finitely many categories such as high, medium, and low. Then, she may switch from one brand to another only if her belief about a brand’s quality moves from one category to another. Although memory limitations in the form of finite memory naturally arise in many facets of economic life, formal economic models that incorporate the finite memory restriction are rather limited. The aim of this project is to understand the implications of finite memory for dynamic decision problems in both choice theoretical and strategic economic contexts. These implications are significant to reach out a wide range of objectives such as obtaining better predictions of consumer’s choices, unravelling the observed equilibrium behaviour, and regulating markets as to maximize social welfare. In addition, the results to be obtained can provide new explanations for observed economic phenomena such as limited attention and price stickiness.The outgoing phase of this GF project is divided into two parts. This first part of the project aims at addressing a fundamental problem to expedite the use of finite memory model in economic applications. In the second part, I aim to examine how strategic pricing aspects of a dynamic duopoly (oligopoly) model. The goal in here is to incorporate memory limitations in the form of finite memory into a fundamental economic problem, and identify its effects.This GF project will be carried out in Princeton University under the supervision of Prof. Faruk Gul. I will then return to Bilkent University to improve my results for 12 months under the supervision of Prof. Semih Koray.
Оригинален текст от CORDIS (на английски).
Участници
- BILKENT UNIVERSITESI VAKIF · Bilkent AnkaraКоординаторТурция
- TRUSTEES OF PRINCETON UNIVERSITY · Princeton, NjСъединени щати
Връзки
Данни: CORDIS, © Европейски съюз