Abstract: A fault risk early warning method of distribution network based on improved RelieF-Softmax algorithm is proposed. Four categories including 24 fault features of distribution network are determined through data investigation and preprocessing. Considering the frequency of distribution network faults and their consequences, the risk classification method of distribution network is presented. K-maxmin clustering algorithm is introduced to improve the random sampling process, and an improved RelieF feature extraction method is proposed to obtain the optimal feature subset with the strongest correlation and minimum redundancy. The loss function of Softmax is improved to cope with the influence of sample imbalance on the prediction accuracy. The optimal feature subset and Softmax classifier are applied to forewarn the fault risk of distribution network. The fault risks of 191 feeders in the south of China are warned, and the test results demonstrate the effectiveness of the proposed method.