Regional evaluation of debris-flow danger degree is an important part of environment protection and disaster reduction.Based on immune evolutionary algorithm and by collecting typical and complete samples,the projection pursuit cluster model(PPC) is established and applied to treat the multi-factor evaluation of danger degree of debris flow.Projection pursuit can project high dimensional data to low dimensional space.Through studying the main characteristics in the 1-D space,the key information in original data can be obtained,which can not only eliminate the subjectivity of the weight of each factor,but also reveal its impacts on the evaluation according to its actual mean contribution rate,based on the classification of different regional danger degrees of debris flow.In this way,the key factors can be clearly selected.The study further deepens the quantitative understanding of the evaluation factors determining regional danger degree of debris flow.