Abstract
Uncertainty and ambiguity are frequently involved in the decision-making process in our daily life. This paper develops a generalized decision-making framework based on the prospect theory under an intuitionistic fuzzy environment, by closely integrating the prospect theory and the intuitionistic fuzzy sets into our framework. We demonstrate how to compute the intuitionistic fuzzy prospect values as the reference values for decision-making and elaborate a four-step editing phase and a valuation phase with two key functions: the value function and the weighting function. We then conduct experiments to test our decision-making methodology and the key features of our framework. The experimental results show that the shapes of the value function and the weighting function in our framework are in line with those of prospect theory. The methodology proposed in this paper to elicit prospects that are not only under uncertainty but also under ambiguity. We reveal the decision-making behavior pattern through comparing the parameters. People are less risk averse when making decisions under an intuitionistic fuzzy environment than under uncertainty. People still underestimate the probability of the events in our experiment. Further, the choices of participants in the experiments are consistent with the addition and multiplication principles of our framework.
Introduction
Traditional models of rational choice, such as Expected Utility Theory (EUT), often fail to predict actual human behavior because they do not account for "framing effects" or the asymmetric treatment of gains and losses. While Kahneman and Tversky’s Prospect Theory (PT) addressed these issues by introducing subjective value and weighting functions, it typically assumes crisp, clear probabilities. In reality, decision-makers often face information that is "fuzzy" or ambiguous (e.g., "about a 30% chance").
This research introduces a generalized framework that merges PT with Intuitionistic Fuzzy Sets (IFS). Unlike standard fuzzy sets, IFS accounts for three distinct states: membership (pro), non-membership (con), and hesitancy (indeterminacy). This allows for a more robust mathematical model capable of handling the inherent ambiguity of real-world data.
Methodology
The proposed framework consists of two primary phases:
1. The Editing Phase
Raw, ambiguous information is organized through four distinct steps:
- Coding: Establishing a reference point to distinguish "gains" from "losses."
- Combination: Aggregating parallel or sequential events using specific fuzzy operators (addition and multiplication principles).
- Scoring: Converting fuzzy information into a "score function" ($s = \mu - \sigma$), representing the net certainty level.
- Correction: Smoothing extreme or overly complex values to simplify the decision space.
2. The Valuation Phase
The core of the model calculates the Intuitionistic Fuzzy Prospect Value (IFPV) using two modified functions:
- Modified Value Function: Utilizes Interval-Valued Intuitionistic Fuzzy Values (IVIFV) to handle vagueness in monetary outcomes (e.g., "winning approximately $100").
- Modified Weighting Function: Adapts the probability weighting function to use the fuzzy "score" as the independent variable, including a "penalty" term to account for the psychological impact of hesitancy.
Experimental Results and Analysis
The framework was validated through controlled experiments involving 12 students from Sichuan University. Participants evaluated 80 double-outcome gambles across various levels of fuzzy scores to fit the parameters of the value and weighting functions.
Key Findings
- Ambiguity Optimism: Surprisingly, the experiments revealed that individuals are less risk-averse under intuitionistic fuzzy environments than under pure uncertainty. When faced with indeterminacy, participants tended to be over-optimistic, effectively assigning the unknown portion of the probability to the positive outcome.
- Parameter Shifts: The risk attitude parameter ($\alpha$) increased from 0.82 (typical in standard PT) to 0.90. While $\alpha < 1$ still indicates risk aversion, the shift toward 1 suggests that fuzzy ambiguity pushes decision-makers toward more risk-neutral behavior.
- Weighting Function Consistency: As shown in Figure 3b, the weighting function maintains the characteristic "inverse-S-shape" of Prospect Theory. This confirms that even in fuzzy environments, people continue to overestimate low-probability events and underestimate high-probability ones.
Conclusion
This framework successfully generalizes Prospect Theory to accommodate intuitionistic fuzzy information, making standard PT a special case where the degree of indeterminacy is zero. While the study is limited by a small sample size and the use of simple monetary gambles, it provides a foundation for applying fuzzy logic to complex real-world decision-making in fields like engineering and public policy. Future work will focus on refining the estimation of fuzzy penalty terms and expanding the application scope to multi-attribute decision-making.
Summary updated