Portfolio Optimization of Uncertain Random Returns based on Partial Exponential Entropy

Abstract

Partial entropy is a device to measure uncertainty of an uncertain random variable. In order to characterize indeterminacy of uncertain random variables, the concept of partial exponential entropy for uncertain random variables is presented. By invoking inverse uncertainty distribution, we derive a formula for computing partial exponential entropy for uncertain random variables. As an application of partial exponential entropy, portfolio selection problems for uncertain random returns are optimized via partial entropy-mean models. For better understanding, several examples are provided.