度小滿金融大數據風控模型實踐

{"type":"doc","content":[{"type":"blockquote","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"導讀：金融是AI賦能傳統行業的重要賽道。本次分享的主要內容爲金融大數據風控模型在度小滿金融的實踐。主要介紹金融大數據風控模型的主要技術方法與在應用層面的主要問題，並結合新冠疫情背景下，探討下風控模型的發展。"}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"金融大數據風控模型的技術方法"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"1. 風險管理中的金融科技"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https:\/\/static001.infoq.cn\/resource\/image\/e9\/f7\/e9eea1b3e3ca2cf2fae408c1b55606f7.png","alt":null,"title":"","style":[{"key":"width","value":"75%"},{"key":"bordertype","value":"none"}],"href":"","fromPaste":false,"pastePass":false}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"風險管理中的金融科技主要包括兩個方面："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"一方面是傳統金融的風險管理，包括："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"A卡（Application Scorecard，申請評分卡）"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"B卡（Behavior Scorecard，行爲評分卡）"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"C卡（Collection Scorecard，催收評分卡）"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"三張評分卡，是傳統金融科技中的一部分。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"另一方面是信息技術，包括："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"算法算力的大幅度提升：A（Artificial Intelligence，人工智能）"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"用戶的行爲數據數字化的存儲和挖掘基礎：B（Big Data，大數據）"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"強大的資源服務共享能力：C（Cloud，雲服務）"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"利用信息技術能力能夠有效提升傳統風險管理評分卡建模的效果。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"2. 度小滿信貸風險"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https:\/\/static001.infoq.cn\/resource\/image\/da\/2b\/da602ca1587d76b3015474af4f51482b.png","alt":null,"title":"","style":[{"key":"width","value":"75%"},{"key":"bordertype","value":"none"}],"href":"","fromPaste":false,"pastePass":false}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"在度小滿信貸業務發展過程中，積累了大量數據和模型相關的實戰經驗。下面主要介紹關於信貸風險模型在度小滿的實戰經驗。如何去識別信用風險，其核心關鍵點在於識別借款用戶真實合理的資金需求以及評估用戶是否擁有較好的還款意願和能力，主要包括三個方面："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"底層需要的是用戶基礎畫像信息，主要包括用戶的年齡、性別、學歷、婚姻狀況、職業、收入、消費能力、房車等資產以及相應的歷史信用信息，金融相比於電商等領域對基礎畫像的準確度等更爲嚴格，因其涉及了用戶真實的還款能力。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"中層爲用戶的基礎行爲需求模式，主要是用戶當前資金端行爲往往與前一段時間內的行爲存在較大的相關性，通過這些行爲可以預測用戶的真實資金需求以及未來的還款表現。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"頂層爲用戶的社會活動，即“物以類聚，人以羣分”，通過關聯網絡中的聚集性發現欺詐團伙，同時可以發現與還款意願差關係緊密的人逾期的風險更高，而與還款行爲好的用戶關係密切的人逾期風險相對較低。"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"核心而言，利用用戶的基礎畫像、行爲需求，外延到用戶社會活動去挖掘用戶是否有真實合理的資金需求，是否有較好的還款意願和還款能力，從而建立區分度良好的風險模型。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"3. 時間序列的處理：貸前"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https:\/\/static001.infoq.cn\/resource\/image\/69\/f5\/69d5888242a2dc9bb26b0b5fb6c68cf5.png","alt":null,"title":"","style":[{"key":"width","value":"75%"},{"key":"bordertype","value":"none"}],"href":"","fromPaste":false,"pastePass":false}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"信貸業務通過用戶授權獲取徵信報告，基於徵信報告瞭解用戶的信用歷史，通過分析用戶的行爲時序來理解用戶的真實現金流需求。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"bulletedlist","content":[{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"常見的徵信查詢包括信用卡審批、貸款審批等，此類查詢表徵了用戶在該時刻的資金需求，通過徵信報告中貸款發放情況匹配貸款查詢申請時間，可以分析用戶資金的用信行爲。傳統金融行業常利用諸如基於不同時間滑動窗口的加工邏輯方式去進行統計，包括過去一個月三個月、六個月、十二個月，二十四個月徵信報告查詢次數等、過去一個月、三個月貸款發放筆數等指標。"}]}]},{"type":"listitem","attrs":{"listStyle":null},"content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"徵信報告還包括用戶的公司地址變更、公積金變更等信息，將上述信息基於時間軸進行劃分，就可以對用戶在一段時間內的信貸需求和用信情況進行刻畫分析。我們利用深度神經網絡去進行分析，通過記錄時間點、該時間點的動作、該動作的類型以及該動作的多個特徵值，將每一項信息轉爲Item作爲輸入提供給LSTM單元，獲取該單元的輸出同時又作爲下一個時間序列的輸入，不斷學習從而獲得更好的信息表達。"}]}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"相比傳統方案，基於這類機器學習的方案能夠帶來KS 2個點的提升。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"4. 時間序列的處理：貸中"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"image","attrs":{"src":"https:\/\/static001.infoq.cn\/resource\/image\/9a\/cf\/9a59f8fe5baebc45da9eaf65165795cf.png","alt":null,"title":"","style":[{"key":"width","value":"75%"},{"key":"bordertype","value":"none"}],"href":"","fromPaste":false,"pastePass":false}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}}]}