Method on Determining Feature Weight in Case-Based Reasoning System

At present, the determining method on feature weight have the problems of complexity and one-sidedness of determining in case-based reasoning system. In this paper, an integrated method based on BP Neural Network and sensitivity analysis is studied. An algorithm based on BP Neural Network and sensitivity analysis (BPNN-SA) is put forward, the network topology changes in accordance with input nodes change. A red-seed watermelon case is given to illustrate this determining method. After the calculation and analysis, results show the new method is better than the current methods and effective for determining feature weight.