Development of human decision making model with consideration of human factors through reinforcement learning and prospect utility theory

Abstract

Human decision-making (HDM) is a complex process. Various human factors play a significant role in this process. Human factors directly or indirectly affect the entire process of decision-making (DM). In this study, an attempt has been made to integrate some of the human factors like past experiences (pe), emotion (ef), times factors (tf), and uncertain (un) with the reinforcement learning (RL) method to develop model for HDM. For this Iowa gambling Task (IGT) has been used as a data collection tool, data of 57 subjects were collected. It is a well-known experience-based task that helps to identify the DM behaviour of participants. An AHP method has been also used to decide the criteria weight to different human factors in the HDM model. Four learning models are developed that are the combination of different utility functions, learning rules, and choice rules. The AHP method decides the preference of various factors incorporated in the developed models. From the results, it is observed that the model based on prospect utility, decay RL, and trial dependency (PU-DRI-TDC Model) performs better when the emotion factor was given the highest preference than others. In addition to this, the IGT learning of participants was also analysed.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 5 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.