本文關(guān)鍵詞： 貸款違約預(yù)測(cè) 非平衡數(shù)據(jù) 隨機(jī)森林 并行計(jì)算　出處：《中南大學(xué)》2013年碩士論文　論文類型：學(xué)位論文

【摘要】：如何在發(fā)放貸款前有效的評(píng)價(jià)和識(shí)別借款人潛在違約風(fēng)險(xiǎn),計(jì)算借款人的違約概率,是現(xiàn)代金融機(jī)構(gòu)信用風(fēng)險(xiǎn)管理的基礎(chǔ)和重要環(huán)節(jié),也是數(shù)量經(jīng)濟(jì)學(xué)、金融學(xué)等領(lǐng)域中的研究熱點(diǎn)問(wèn)題。 現(xiàn)有的貸款違約數(shù)據(jù)大部分都是非平衡的,以往的研究并未足夠注意這一特征也就重視這一問(wèn)題的深入研究,本文主要研究如何借助非平衡數(shù)據(jù)分類的思想對(duì)銀行等金融機(jī)構(gòu)的歷史貸款數(shù)據(jù)進(jìn)行分析,預(yù)測(cè)貸款違約的可能性。對(duì)于非平衡數(shù)據(jù)問(wèn)題,采用基于數(shù)據(jù)平衡的方法構(gòu)造隨機(jī)森林；針對(duì)數(shù)據(jù)較大的問(wèn)題,采用具有并行特性的隨機(jī)森林算法�；谏鲜鲅芯�,本文提出了一種改進(jìn)的帶權(quán)重的并行平衡隨機(jī)森林算法(WPBRF)。WPBRF算法在構(gòu)造隨機(jī)森林的每個(gè)決策樹的同時(shí)利用OOB數(shù)據(jù)估計(jì)該決策樹的預(yù)測(cè)性能,并據(jù)此賦予每個(gè)決策樹不同的權(quán)重；此外,WPBRF算法利用了隨機(jī)森林算法的可并行計(jì)算的特點(diǎn),減少了單個(gè)決策樹的訓(xùn)練時(shí)間。 實(shí)驗(yàn)結(jié)果表明,WPBRF在準(zhǔn)確率和平衡準(zhǔn)確率等方面超過(guò)了SVM、KNN、C4.5等常見(jiàn)分類算法和隨機(jī)森林算法。此外,利用隨機(jī)森林的并行性的WPBRF算法大幅降低了算法的學(xué)習(xí)時(shí)間,提高了算法的執(zhí)行效率。
[Abstract]:How to effectively evaluate and identify the borrower's potential default risk and calculate the borrower's default probability before making loans is the foundation and important link of credit risk management of modern financial institutions, and it is also the quantitative economics. Hot issues in finance and other fields. Most of the existing loan default data are unbalanced. Previous studies have not paid enough attention to this feature and paid attention to the in-depth study of the problem. This paper mainly studies how to use the idea of disequilibrium data classification to analyze the historical loan data of banks and other financial institutions to predict the possibility of loan default. A method based on data balance is used to construct a random forest, and a stochastic forest algorithm with parallel characteristics is used to solve the problem of large data. In this paper, an improved parallel balanced stochastic forest algorithm with weight is proposed, which not only constructs each decision tree of random forest, but also uses OOB data to estimate the prediction performance of the decision tree, and gives different weights to each decision tree. In addition, the WPBRF algorithm takes advantage of the parallelism of the stochastic forest algorithm to reduce the training time of a single decision tree. The experimental results show that WPBRFs outperform common classification algorithms such as SVMKNNC4.5 and stochastic forest algorithms in terms of accuracy and balance accuracy. In addition, the WPBRF algorithm using the parallelism of stochastic forests greatly reduces the learning time of the algorithm. The efficiency of the algorithm is improved.
【學(xué)位授予單位】：中南大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2013
【分類號(hào)】：F830.59;F224

【參考文獻(xiàn)】

相關(guān)期刊論文 前7條

1 顧乾屏;孫曉昆;陳兵;蔣林宏;聞國(guó)平;;信貸違約率實(shí)證分析[J];北京工商大學(xué)學(xué)報(bào)(自然科學(xué)版);2007年02期

2 彭建剛;易宇;李樟飛;;測(cè)算商業(yè)銀行貸款違約概率的貸款違約表法探討[J];管理學(xué)報(bào);2009年06期

3 劉鐵牛;;商業(yè)銀行貸款違約模型研究綜述[J];湖南商學(xué)院學(xué)報(bào);2009年03期

4 夏紅芳;馬俊海;;基于KMV模型的上市公司信用風(fēng)險(xiǎn)預(yù)測(cè)[J];預(yù)測(cè);2008年06期

5 梁世棟,郭N，

本文編號(hào)：1553824