Rough-Fuzzy Rule Interpolation for Data-Driven Decision Making

Any practical decision making strategy is required to ensure that the best decision is made with respect to the information available and the knowledge possessed by experts. A rule-based fuzzy decision making system typically works on the fuzzy rules generated from numerical data acquired in the problem domain. However, different expert opinions on fuzzy partitions may result in a range of uncertainties in representing the domain knowledge. The invention of rough-fuzzy sets offers a great potential in the representation, handling and utilisation of different levels of uncertainty in knowledge. Inspired by this observation, a rough-fuzzy rule interpolation method is introduced in this paper to enable decision making systems modelling and harnessing additional uncertain information, in order to implement a fuzzy reasoning system that can work with incomplete rule base. An initial experimental investigation is carried out and the results are presented to demonstrate the effectiveness of the proposed method in aiding the development of an intelligent decision making system.