A RAGA-PPC (Real coding based Accelerating Genetic Algorithm-Projection Pursuit Classification) model was employed to convert high dimensional data to low dimensional data and assess the health of a river in Haihe river basin. Majia River, a key river of Haihe river basin in Shandong Province, was taken as case study. Data obtained from water quality monitoring and pollution investigation in 2009-2011 were utilized, and 12 sections were selected for the investigation. The assessment results show that the overall trend of total nitrogen and ammonia nitrogen in the whole river is growing, and also biochemical oxygen demand in the downstream area from Dadaowang gate to Dengji goes almost straight up. Chemical oxygen demand and biochemical oxygen demand of the river change in flood season and non-flood season, both increasing in flood season. According to comparison between the weighted value and the standard value of 5 sections, the chemical oxygen demand and biochemical oxygen demand are both between 0.22-0.15, indicating that Majia River is basically in a sub-healthy state. Furthermore, the human factors and natural factors of Majia River's sub-healthy state are analyzed in the purpose of providing effective scientific basis for river health assessment and river pollution control in the Haihe River Basin.