Stochastic Decision Tree Acceptability Analysis Incorporating Attribute Interactions

In rapidly evolving decision environments, forecasting future states poses a significant challenge. Under conditions of uncertainty, traditional decision-making approaches struggle to address unknown or uncertain probabilities of environmental states. The Stochastic Decision Tree Acceptability Analysis (SDTAA) method is specifically designed to tackle decision problems in scenarios where environmental state probabilities are unknown. It models these probabilities using a state probability space and ranks alternatives based on their rank acceptability and expected value. While SDTAA proves highly effective, it assumes attribute independence, overlooking potential positive or negative interactions between attributesinteractions that are prevalent in real-world decision problems and significantly influence outcomes. To address this limitation, this study integrates the Choquet integral into the traditional SDTAA framework, proposing the Stochastic Decision Tree Acceptability Analysis with Attribute Interactions (SDTAI). This approach effectively captures complex attribute interactions while preserving SDTAA's capacity to model environmental uncertainty. Additionally, to operationalize the proposed SDTAI method, we develop an algorithm based on Monte Carlo simulation and validate its efficacy through a case study on startup investment decisions.