A Fuzzy Decision Tree Based Approach for the Selection of Investments

Decision trees have been widely exploited in investment analysis for projects selection based on specified criteria such as Net Present Value (NPV). Since such framework is initially established upon a precise information basis, it fails in providing reliable decisions when uncertainty is present at the level of the project data. To cope with such shortcoming, we propose a Fuzzy Net Present Value (FNPV) measure by fuzzification of the elementary project parameters within a fuzzy decision tree structure. In this context, we introduce the concept of entropy as a selection criterion of the best alternative. A genetic algorithm is investigated to determine lower and upper bounds associated to <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\alpha$</tex>-cuts of the fuzzy measures. For the evaluation of the proposed framework, an industrial plant has been considered as a case study.