An efficient algorithm for multisensory data fusion under uncertainty condition

This paper addresses multisensory data fusion for unknown systems. The main focus is on identifying and dealing with uncertainty and inconsistent conditions. Most data fusion methods depends on system behavior, which do not allow to easily deal with unknown systems. This method, works based on a clustering technique followed by an MLP predictor. It is specifically designed for unknown systems in uncertain and inconsistent conditions. However, it can also be applied for known and exact sources. When the sources contain uncertainty and inconsistency, data fusion may fail. The proposed method can recognize and remove the inconsistent data points and produces reliable results. The experimental results on both synthetic and real data confirm the effectiveness of the proposed approach.