Interactive portfolio optimization with cognition-limited human decision making assisted by auxiliary factors
crossref(2022)
Abstract
Abstract The success of an interactive method of portfolio optimization depends on not only the acquisition of a satisfactory solution for a decision maker (DM) but also the amount of cognitive effort the DM affords during interactions. For the purposes of implementing a successful interactive method, however, it is often unrealistic to assume a participating DM with a high cognitive capability. Confronted with the challenge of how a cognition-limited DM is able to remove uncertainties in his/her preference feedback and make more confident judgments during interactions so that a stable, satisfactory solution is found as early as possible, we conduct innovative investigations with respect to DM preference articulation. First, in addition to the objective functions of an optimization model, which are used in the present interactive methods as the only references for DM preference articulation, we aim to define the candidate auxiliary factors involved in the portfolio optimization model and recognize those that are closely correlated with the objective functions (called primary factors) and that can help a DM differentiate solutions if he/she is uncertain in making a judgment according to the primary factors. Next, in order to define an artificial DM to simulate the DM’s decision behaviors being assisted by auxiliary factors in the interactive process, based on a deliberately formulated value function representing the DM’s true preferences on portfolios, we introduce the decision strength (DS) to measure the DM’s capability of making a selection between two solutions. Thus, the DM’s decision behaviors are reflected by his/her DS changing with an increasing number of auxiliary factors incorporated into the primary factors. To verify the proposed interactive method for portfolio optimization, we conduct extensive comparative experiments under the participation of three types of artificial DMs with different levels of cognitive capabilities, comparative experimental results with and without auxiliary factors with respect to four performance measures are presented.
MoreTranslated text
AI Read Science
Must-Reading Tree
Example
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined