With　the　development　of　the　telecommunications　industry,　the　business　products　of　major　operators　continue　to　innovate,　and　it　is　particularly　important　to　analyze　the　credit　value　of　telecom　users　and　use　them　for　business　risk　control　and　management.　Based　on　the　historical　behavior　data　of　telecom　users,　based　on　the　traditional　Logistic　regression　model　and　the　support　vector　machine　model,　this　paper　proposes　a　risk　prediction　model　combining　the　two　methods,　trying　to　find　effective　measures　to　reduce　the　risk　of　telecom　users.　The　experimental　results　show　that　compared　with　the　two　single　models,　the　combined　prediction　model　not　only　has　higher　classification　accuracy,　but　also　obtains　better　robustness,　which　can　effectively　predict　the　risk　of　telecom　users.　©　2019　IEEE.