本研究主要針對「信用紀錄薄弱」之社會新鮮人之資金借貸進行探討。本文發現，社會新鮮人有收入與支出不對等的問題，倘若遇到臨時性資金需求時，有向銀行貸款的潛在需求。然而社會新鮮人信用紀錄相對薄弱，現行銀行體系無法有效改善此一問題。
本文提出針對社會新鮮人之P2P模式的微型信貸平台，有望改善此一問題。此一微型信貸平台能夠先行蒐集社會新鮮人之借貸資料，日後當社會新鮮人有大額貸款需求，向銀行等金融機構借款時，銀行可回頭參考社會新鮮人在平台所留下的信貸紀錄，更能精確衡量其潛在違約風險，減少向銀行借貸實相關資訊的缺漏。此外，銀行更能直接與P2P平台合作，以達到互利共生，有效擴大信貸業務至過去無法觸及的族群，能夠擴張在個人金融借貸的業務發展與整體健全性。
本研究使用機器學習方法建構是否為準時償還貸款預測模型，期望能作為日後銀行的參考。根據實證結果發現，隨機森林模型、自適應增強模型以及極限梯度提升模型，三者預測整體準確率在測試資料中皆可達到至少77%的準確率，AUC值也皆落在約0.7，都比單純只使用決策樹模型能得到更好的效果。同時此三模型在單純只預測準時償還的貸款是更加準確的，在特異率與陰性預測比率皆有相當高的水準。

This study mainly focus on the newly graduated students, borrowing needs. Newly graduated students face the problem that income and expenditure are not balanced. They may have potential demands of borrowing from banks. However, newly graduated studnets only have limited credit history that banks heavily reply on when evaluating the borrower. Traditional bank lending system may not be the solution of this issue. This study proposes a newly graduated student P2P micro loan platform to expect solove this problem.
By matching newly graduated studnet as borrower and lender, the micro loan P2P platform collects the newly graduated student credit information. Once newly graduated students have large amount of loan demands from banks afterward, banks can refer back their credit records left on the platform. By working with the plateform, banks can not only effectively evaluate borrowers’ potential default risks but also expand business to previously unreachable clients.
This study uses machine learning method to establish a model to predict whether the loan is fully paid or not, which is expected to serve as a reference for banks. According to the empirical results, random forest model, AdaBoost model and XgBoost model overall accuracy can reach at least 77% and AUC is about 0.7 in the testing data. Three of them is better than using simply decision tree model. Moreover, these three models are more accurate in predicting fully paid loans, the ratio of negative predictive value and specificity are both high.

中文部分
世界銀行(2017)。《2017年全球普惠金融指數資料庫：度量普惠金融和金融科技革命》報告
林容伊、簡琬真、王薏慈 (2017)。專屬大學生借貸平台借貸門檻不設限。檢自https://vita.tw/%E5%B0%88%E5%B1%AC%E5%A4%A7%E5%AD%B8%E7%94%9F%E5%80%9F%E8%B2%B8%E5%B9%B3%E5%8F%B0-%E5%80%9F%E8%B2%B8%E9%96%80%E6%AA%BB%E4%B8%8D%E8%A8%AD%E9%99%90-7ac0546ce3aa (Jun. 12, 2018)
張小玫、周意庭 (2016)。Fintech金融科技新創公司(二)－SoFi。檢自http://iknow.stpi.narl.org.tw/post/Read.aspx?PostID=12769 (Jun. 9, 2018)
經濟部中小企業處 (Apr. 5, 2011)。檢自https://friap.moeasmea.gov.tw/kn_article.php?nid=111&gid=2 (Jun. 2, 2018)

英文部分
Arner, D. W., Barberis, J. and Buckley, R. P. (2016). The evolution of Fintech: A new post-crisis paradigm. Georgetown Journal of International Law, 47(4), 1271-1319.
Breiman, L. (1996). Bagging predictors. Machine Learning, 24(2), 123–140.
Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5-32.
Breiman, L., Friedman, J. H., Stone, C. J., and Olshen, R.A. (1984). Classification and regression trees, Taylor & Francis.
Choy, S. P., and Li, X. (2006). Dealing with debt: 1992–93 bachelor’s degree recipients 10 years later (NCES 2006-156). Washington, DC: US Department of Education, National Center for Education Statistics.
Christman, D. E. (2000). Multiple realities: Characteristics of loan defaulters at a two-year public institution. Community College Review, 27(4), 16-32.
Chen, T. and Guestrin, C. (2016). XGBoost: A scalable tree boosting system, KDD '16 Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785-794
Collins, F. S., and Varmus, H. (2015). A new initiative on precision medicine. New England Journal of Medicine, 372(9), 793-795.
Cull, R., Demirgu¨c¸-Kunt, A., and Morduch, J. (2009). Microfinance meets the market. Journal of Economic Perspectives, 23(1), 167-192.
Dynarski, M. (1994). Who defaults on student loans? Findings from the national postsecondary student aid study. Economics of Education Review, 13(1), 55-68.
Flint, T. A. (1994). The federal student loan default cohort: A case study. Journal of Student Financial Aid, 24(1), 13-30.
Flint, T. A. (1997). Predicting student loan defaults. The Journal of Higher Education, 68(3), 322-354.
Friedman, J. (2001). Greedy function approximation: A gradient boosting machine. The Annals of Statistics, 29(5), 1189-1232.
Greene, L. L. (1989). An economic analysis of student loan default. Educational Evaluation and Policy Analysis, 11(1), 61-68.
Harrast, S. A. (2004). Undergraduate borrowing: A study of debtor students and their ability to retire undergraduate loans. Journal of Student Financial Aid, 34(1), 21-37.
Herr, E., and Burt, L. (2005). Predicting student loan default for the University of Texas at Austin. Journal of Student Financial Aid, 35(2), 27-49.
Ho, T. K. (1995). Random decision forest. Proceeding of the 3rd International Conference on Document Analysis and Recognition, Montreal, Canada, 278-282
Ho, T. K. (1998). The random subspace method for constructing decision forests. Pattern Analysis and Machine Intelligence, 20(8), 832-844.
Jacob, P. K., Cekic, O., Hossler, D., and Hillman, N. (2009). What Matters in Student Loan Default: A Review of the Research Literature. Journal of Student Financial Aid, 39(1), 19-29.
Knapp, L. G., and Seaks, T. G. (1992). An analysis of the probability of default on federally guaranteed student loans. The Review of Economics and Statistics, 74(3), 404-411.
Li, Y. (2017). Risk management of P2P internet financing service platform. Paper presented at the IOP conference series: Materials science and engineering.
Lochner, L., and Monge-Naranjo, A. (2004). Education and default incentives with government student loan programs. Cambridge, MA: National Bureau of Economic Research.
Mateescu, A. (2015). Peer-to-Peer Lending. Data and Society, 1-23
Podgursky, M., Ehlert, M., Monroe, R., Watson, D., and Wittstruck, J. (2002). Student loan defaults and enrollment persistence. Journal of Student Financial Aid, 32(3), 27-42.
Schwartz, S., and Finnie, R. (2002). Student loans in Canada: An analysis of borrowing and repayment. Economics of Education Review, 21(5), 497-512.
Seifert, C. F., and Worden, L. (2004). Two studies assessing the effectiveness of early intervention on the default behavior of student loan borrowers. Journal of Student Financial Aid, 34(3), 41-52.
Singh, S. and Gupta, P. (2014), Comparative study Id3, cart and C4.5 decision tree algorithm: A Survey. International Journal of Advanced Information Science and Technology, 3(7), 47-52.
Volkwein, J. F., and Szelest, B. P. (1995). Individual and campus characteristics associated with student loan default. Research in Higher Education, 36(1), 41-72.
Wilms, W. W., Moore, R. W., and Bolus, R. E. (1987). Whose fault is default? A study of the impact of student characteristics and institutional practices on guaranteed student loan default rates in California. Educational Evaluation and Policy Analysis, 9(1), 41-54.
Woo, J. H. (2002). Factors affecting the probability of default: Student loans in California. Journal of Student Financial Aid, 32(2), 5-23.

網路連結
Four risks in peer-to-peer lending. Retrieved from https://www.bondora.com/blog/4-risks-in-peer-to-peer-lending/ (May 7, 2018)
Rind, V. (2016). Pros and cons of peer-to-peer lending. Retrieved from https://www.gobankingrates.com/loans/personal/5-perks-peer-to-peer-lending/
美國最大消費貸款平台SoFi CEO Mike Cagney：我們有潛力把客戶從銀行手裡搶過來(Jul. 10, 2016)。檢自https://read01.com/GRGNRB.html (Jun. 4, 2018)
為一流歐美商學院MBA留學生提供助學貸款的Prodigy Finance獲1億英鎊融資(Aug. 10, 2015)。檢自https://36kr.com/p/5036265.html (Jun. 18, 2018)
讀MBA，向校友借錢 (Sep. 25, 2017)。檢自https://read01.com/GP4A362.html (Jun. 19, 2018)