學術產出-學位論文
文章檢視/開啟
書目匯出
-
題名 基於隨機森林模型下P2P網路借貸違約預測
The prediction of default in P2P Lending based on Random Forest Model作者 吳志龍
Wu, Zhi-Long貢獻者 廖四郎
Liao, Szu-Lang
吳志龍
Wu, Zhi-Long關鍵詞 P2P借貸
隨機森林模型
Logistic回歸模型
個人信用風險評估
P2P Lending
Random Forest
Logistic regreesion
Private credit risk evaluation日期 2019 上傳時間 7-八月-2019 16:13:06 (UTC+8) 摘要 摘要\n本研究使用傳統的Logistic回歸模型與機器學習的隨機森林模型對P2P借貸的個人信用風險進行評估預測。本研究的數據來源於LendingClub的2018年度公開數據資料，先對P2P借貸的個人信用風險因素進行挑選，再使用挑選出的變量對Logistic回歸模型與隨機森林模型進行訓練，並用測試集檢驗兩個模型對個人信用風險的預測能力。結果表明，隨機森林模型的在決策樹為800棵，每棵決策樹的特征值為3個的時候，隨機森林模型預測準確率最高。與Logistic回歸模型比較，隨機森林模型有著更高的精度。本研究還對兩個模型進行了模型性能的比較，結果表明隨機森林的模型性能好過去Logistic回歸模型。\n\n關鍵詞：P2P借貸、隨機森林模型、Logistic回歸模型、個人信用風險評估
Abstract\nIn this paper we evaluate and predict the private credit risk in P2P lending by using traditional Logistic regression and Random Forest in machine learning. For the open data from LendingClub in 2018, we select the private credit risk factors of P2P lending first, and train the Logistic regression and Random Forest model with selected variables. We test the prediction ability of two models with test set. The result shows that when there are 800 decision trees and 3 features for each tree in Random Forest model, accuracy of the model reaches best. Compared with Logistic regression, Random Forest has higher precision. We also compare the performance of two models, which shows that Random Forest is better than Logistic regression.\n\nKeywords: P2P Lending, Random Forest, Logistic Regression, Private Credit Risk Evaluation參考文獻 Bekhet,H.A.,Eletter,S.F.K.(2014).Credit risk assessment model for Jordanian commercial banks: Neural scoring approach. Review of Development Finance, Vol.4,20-28.\nBerkson,J.(1944). Application of the Logistic Function to Bio-Assay. Journal of the American Statistical Association , Vol.39, 357-365.\nBreiman,L.(2001). Random forest. Machine learning, 45(1), 5-32\nBruett,T.(2007).Cows, Kiva, and Prosper. Com: how disintermediation and the Inter net are changing microfinance. Community Development Investment Review, 2, 44-50.\nCaemichael,D.(2014). Modeling default for peer-to-peer loans . Working Paper.\nCoffman,J.Y.(1986). The proper role of tree analysis in forecasting the risk behavior of borrowers. Management Decision System, No.3, 47-59.\nDasarathy,B.V. Sheela,B.V.(1979). A composite classifier system design: Concepts and methodology. Proceedings of the IEEE, 67(5),708-713.\nEmekter,R.Tu,Y.Jirasakuldech,B&Lu,M.(2015).Evaluating credit risk and loan performance in online Peer-to-Peer(P2P) lending. Applied Economics, 47(1), 54-70\nEverett,C.R.(2015).Group membership, relationship banking and loan default risk : the case of online social lending. Banking and Fiance Review,7(2),14-54.\nFreedman,S.Jin,G.Z.(2008).Dynamic Learning and Selection: the Early Years. University of Maryland.\nGarman,S.R.Hampshire,R.C.Krishnan,R.(2008).Person-to-Person Lending: The Pursuit of (More) Competitive Credit Markets. E-Life: Web-Enabled Convergence of Commerce, Vol.14,54-58\nHerzenstein,M.Andrews,R.L. Dholakia,U.M.Lyandres,E.(2008). The democratization of personal consumer loans Determinants of success in online peer-to-peer lending communities. Boston University School of Management Research Paper.\nJin,G.Z.Freedman,S.(2008). Do Social Networks Solve Information Problem for Peer-to-Peer Lending? Evidence from Prosper.Com.NET Institute Working Paper, 8-43\nKlafft,M.(2008).Peer to peer lending: auctioning microcredits over the internet. Working Paper.\nMaddala,G.S.(1983).Limited-dependent and qualitative variables in econometrics. Cambridge University Press.\nMagee,J.R.(2011).Peer-to-peer lending in the United States: surviving after Dodd-Frank. North Carolina Banking Institute Journal, Vol.15, 139-174\nMakowski,P.(1985). Credit scoring branches out. Credit World,Vol.75, 30-37.\nMateescu,A.(2015).Peer-to-Peer Lending. Data&Society,1-23.\nIyer,R.Khwaja,A.I.Luttmer,E.F.P.&Shue,K.(2009). Screening in new credit market: Can individual lenders infer borrower creditworthiness in peer-to-peer lending?. Working Paper\nOhlson,J.A.(1980). Financial ratios and the probabilistic prediction of bankruptcy. Journal of accounting research, Vol.18,No.1,109-131.\nPope, Sydnor,2011.What’s in a picture, vidence of discrimination from Prosper.com. Journal of Human Resource,Vol.46,No.1, 53-92\nSerrano-cinca,C.Gutierrez-Nieto,B.&Lopez-Palacios,L.(2015).Determinants of default in P2P lending. PloS one, 10(10),e0139427\nStiglitz,J.E.Weiss,A.(1981). Credit rationing in markets with imperfect information. The American economic review, 71(3), 393-410\n王磊、范超、解明明(2014)。數據挖掘模型在小企業主信用評分領域的應用，統計研究，31卷，第10期，89-98。\n張培強(2011)。信用卡客戶的分類研究，生產力研究，第4期， 87-88。\n張萬軍(2015)。基於大數據的個人信用風險評估模型研究，對外經濟貿易大學。\n彭康(2018)。基於P2P網絡借貸的個人信用風險評估，暨南大學。\n蘇杭西子(2018)。基於隨機森林模型的個人信用風險評估研究，湖南大學。\n劉暢(2015)。基於Logistic的P2P網絡借貸信用風險測度研究，安徽財經大學。 描述 碩士
國立政治大學
金融學系
106352043資料來源 http://thesis.lib.nccu.edu.tw/record/#G0106352043 資料類型 thesis dc.contributor.advisor 廖四郎 zh_TW dc.contributor.advisor Liao, Szu-Lang en_US dc.contributor.author (作者) 吳志龍 zh_TW dc.contributor.author (作者) Wu, Zhi-Long en_US dc.creator (作者) 吳志龍 zh_TW dc.creator (作者) Wu, Zhi-Long en_US dc.date (日期) 2019 en_US dc.date.accessioned 7-八月-2019 16:13:06 (UTC+8) - dc.date.available 7-八月-2019 16:13:06 (UTC+8) - dc.date.issued (上傳時間) 7-八月-2019 16:13:06 (UTC+8) - dc.identifier (其他 識別碼) G0106352043 en_US dc.identifier.uri (URI) https://ah.lib.nccu.edu.tw/item?item_id=143203 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 金融學系 zh_TW dc.description (描述) 106352043 zh_TW dc.description.abstract (摘要) 摘要\n本研究使用傳統的Logistic回歸模型與機器學習的隨機森林模型對P2P借貸的個人信用風險進行評估預測。本研究的數據來源於LendingClub的2018年度公開數據資料，先對P2P借貸的個人信用風險因素進行挑選，再使用挑選出的變量對Logistic回歸模型與隨機森林模型進行訓練，並用測試集檢驗兩個模型對個人信用風險的預測能力。結果表明，隨機森林模型的在決策樹為800棵，每棵決策樹的特征值為3個的時候，隨機森林模型預測準確率最高。與Logistic回歸模型比較，隨機森林模型有著更高的精度。本研究還對兩個模型進行了模型性能的比較，結果表明隨機森林的模型性能好過去Logistic回歸模型。\n\n關鍵詞：P2P借貸、隨機森林模型、Logistic回歸模型、個人信用風險評估 zh_TW dc.description.abstract (摘要) Abstract\nIn this paper we evaluate and predict the private credit risk in P2P lending by using traditional Logistic regression and Random Forest in machine learning. For the open data from LendingClub in 2018, we select the private credit risk factors of P2P lending first, and train the Logistic regression and Random Forest model with selected variables. We test the prediction ability of two models with test set. The result shows that when there are 800 decision trees and 3 features for each tree in Random Forest model, accuracy of the model reaches best. Compared with Logistic regression, Random Forest has higher precision. We also compare the performance of two models, which shows that Random Forest is better than Logistic regression.\n\nKeywords: P2P Lending, Random Forest, Logistic Regression, Private Credit Risk Evaluation en_US dc.description.tableofcontents 目錄\n\n第一章 緒論 1\n第一節 研究動機 1\n第二節 研究目的 3\n第三節 研究架構 4\n第二章 文獻探討 5\n第一節 P2P介紹 5\n第二節 個人信用評估因素 6\n第三節 Logistic回歸模型違約預測 7\n第四節 決策樹與隨機森林模型 8\n第三章 研究方法 10\n第一節 隨機森林模型 10\n1.決策樹算法介紹 10\n2.特徵選擇 11\n3.CART演算法 11\n4.決策樹的剪枝 12\n5.決策樹的缺點 13\n6.集成學習 13\n7.隨機森林模型 15\n第二節 Logistic回歸模型 17\n第四章 實證結果 19\n第一節 資料來源 19\n第二節 變數選取 19\n第三節 資料處理 21\n1.缺失值得處理 21\n2.量化處理 21\n3.相關性檢驗 24\n4.資料劃分 26\n第四節 混淆矩陣 26\n第五節 隨機森林模型結果 27\n1.決策樹個數的選擇 27\n2.特徵數量的選擇 28\n3.隨機森林模型預測結果 29\n第六節 Logistic回歸模型預測結果 30\n第七節 不平衡問題處理與結果比較 31\n第八節 模型性能檢測 33\n第五章 研究結果與未來展望 37\n第一節 研究結果 37\n第二節 未來展望 38\n參考文獻 39 zh_TW dc.format.extent 1101860 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0106352043 en_US dc.subject (關鍵詞) P2P借貸 zh_TW dc.subject (關鍵詞) 隨機森林模型 zh_TW dc.subject (關鍵詞) Logistic回歸模型 zh_TW dc.subject (關鍵詞) 個人信用風險評估 zh_TW dc.subject (關鍵詞) P2P Lending en_US dc.subject (關鍵詞) Random Forest en_US dc.subject (關鍵詞) Logistic regreesion en_US dc.subject (關鍵詞) Private credit risk evaluation en_US dc.title (題名) 基於隨機森林模型下P2P網路借貸違約預測 zh_TW dc.title (題名) The prediction of default in P2P Lending based on Random Forest Model en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) Bekhet,H.A.,Eletter,S.F.K.(2014).Credit risk assessment model for Jordanian commercial banks: Neural scoring approach. Review of Development Finance, Vol.4,20-28.\nBerkson,J.(1944). Application of the Logistic Function to Bio-Assay. Journal of the American Statistical Association , Vol.39, 357-365.\nBreiman,L.(2001). Random forest. Machine learning, 45(1), 5-32\nBruett,T.(2007).Cows, Kiva, and Prosper. Com: how disintermediation and the Inter net are changing microfinance. Community Development Investment Review, 2, 44-50.\nCaemichael,D.(2014). Modeling default for peer-to-peer loans . Working Paper.\nCoffman,J.Y.(1986). The proper role of tree analysis in forecasting the risk behavior of borrowers. Management Decision System, No.3, 47-59.\nDasarathy,B.V. Sheela,B.V.(1979). A composite classifier system design: Concepts and methodology. Proceedings of the IEEE, 67(5),708-713.\nEmekter,R.Tu,Y.Jirasakuldech,B&Lu,M.(2015).Evaluating credit risk and loan performance in online Peer-to-Peer(P2P) lending. Applied Economics, 47(1), 54-70\nEverett,C.R.(2015).Group membership, relationship banking and loan default risk : the case of online social lending. Banking and Fiance Review,7(2),14-54.\nFreedman,S.Jin,G.Z.(2008).Dynamic Learning and Selection: the Early Years. University of Maryland.\nGarman,S.R.Hampshire,R.C.Krishnan,R.(2008).Person-to-Person Lending: The Pursuit of (More) Competitive Credit Markets. E-Life: Web-Enabled Convergence of Commerce, Vol.14,54-58\nHerzenstein,M.Andrews,R.L. Dholakia,U.M.Lyandres,E.(2008). The democratization of personal consumer loans Determinants of success in online peer-to-peer lending communities. Boston University School of Management Research Paper.\nJin,G.Z.Freedman,S.(2008). Do Social Networks Solve Information Problem for Peer-to-Peer Lending? Evidence from Prosper.Com.NET Institute Working Paper, 8-43\nKlafft,M.(2008).Peer to peer lending: auctioning microcredits over the internet. Working Paper.\nMaddala,G.S.(1983).Limited-dependent and qualitative variables in econometrics. Cambridge University Press.\nMagee,J.R.(2011).Peer-to-peer lending in the United States: surviving after Dodd-Frank. North Carolina Banking Institute Journal, Vol.15, 139-174\nMakowski,P.(1985). Credit scoring branches out. Credit World,Vol.75, 30-37.\nMateescu,A.(2015).Peer-to-Peer Lending. Data&Society,1-23.\nIyer,R.Khwaja,A.I.Luttmer,E.F.P.&Shue,K.(2009). Screening in new credit market: Can individual lenders infer borrower creditworthiness in peer-to-peer lending?. Working Paper\nOhlson,J.A.(1980). Financial ratios and the probabilistic prediction of bankruptcy. Journal of accounting research, Vol.18,No.1,109-131.\nPope, Sydnor,2011.What’s in a picture, vidence of discrimination from Prosper.com. Journal of Human Resource,Vol.46,No.1, 53-92\nSerrano-cinca,C.Gutierrez-Nieto,B.&Lopez-Palacios,L.(2015).Determinants of default in P2P lending. PloS one, 10(10),e0139427\nStiglitz,J.E.Weiss,A.(1981). Credit rationing in markets with imperfect information. The American economic review, 71(3), 393-410\n王磊、范超、解明明(2014)。數據挖掘模型在小企業主信用評分領域的應用，統計研究，31卷，第10期，89-98。\n張培強(2011)。信用卡客戶的分類研究，生產力研究，第4期， 87-88。\n張萬軍(2015)。基於大數據的個人信用風險評估模型研究，對外經濟貿易大學。\n彭康(2018)。基於P2P網絡借貸的個人信用風險評估，暨南大學。\n蘇杭西子(2018)。基於隨機森林模型的個人信用風險評估研究，湖南大學。\n劉暢(2015)。基於Logistic的P2P網絡借貸信用風險測度研究，安徽財經大學。 zh_TW dc.identifier.doi (DOI) 10.6814/NCCU201900322 en_US
