A mass assignment algorithm for Dempster-Shafer approach

In this work, a new methodology is proposed for probability mass assignment to be used in Dempster-Shafer approach. The approach is developed for the cases where the prior and the current information are represented by probability density functions. Masses are assigned by comparing the prior and the current probability density functions which are assumed to be Gaussian. The proposed method is tested by artificial data and compared with the method given in [13] which is most similar to the proposed one in the literature.