An Artificial Intelligence Approach for Decision Making in Investment

This paper considers decision making for fuzzy investments. A proposed client asset allocation model based on Mamdani FIS. The major aim of building such model is to help advising clients how to allocate portions of their investments in three types of assets, saving account, investment certificate and investment fund. The investment advisory models are applied with the goal of maximizing the expected return under two constraints, client’s risk and age. The investment decisions can be undertaken when the aggregated if-then rules are applied in Mamdani FIS. A comparison between using different types and number of membership functions is outlined.The types of membership functions used are triangular MF ,Trapizodial MF and mixing between both triangular an trapizodial MF. It was found that using the MFs of the same kind triangular MF only or trapezoidal MF only give better expected returns than mixed MFs. The work is accompanied by an illustrative case study that show the validity of the approach, followed by some recommendations for future research area.