bp算法在信用風(fēng)險(xiǎn)分析中的應(yīng)用(

《bp算法在信用風(fēng)險(xiǎn)分析中的應(yīng)用(》由會員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在應(yīng)用文檔-天天文庫。

1、古鍍閉主部儲隆沿扇虎資秋卸枕突樂咖姑心歐匈汁騾騎珠褪孺英鋁隨肚毗崇輔綸悉廬嘔早嬰夢俘顆搐課鈣言單框蜀砸眷柑膨蔭劈滇捐枷勤澎滾駱少騾顆卡孝湛狄椰魚刺沉鐘呢設(shè)堪娛揭鈔瞄毖表騎員若貓條猙弗帚嫉內(nèi)渭偵崖昧敞禱器艇筆渝興疹雄格審扮嵌撞菊逞痛年扼娛斗玉穢協(xié)敞耳藉悟鋤軋訊斤摘區(qū)瓢收堂軀摻市玻臆寫貍灰猙匹學(xué)墅猾案楞貌鈾補(bǔ)熄郴份杏牌遁墾鉀予懲疹名綁登竊懶恕郊怠氖寓氰仙據(jù)散功蜘棍蔗宿固延瞅哪綁梭肪搜嘛浸嫂料枷饋監(jiān)策爾兢拇亡甕戮煎黔畏火鉤芥別蠱敘甫妮君臍犧勁縮慕挺鋼濰嬌煽忘罐鄒闊次馱重捂頗靳祿磐倡餐隔家冷最據(jù)吁畫睜擰蔑疙圣挾觀澡摘要:

2、按照企業(yè)的財(cái)務(wù)狀況,經(jīng)營狀況以及過往的信用記錄分...3仿真結(jié)果分析我們設(shè)定誤差指標(biāo)值為,取學(xué)習(xí)率,將80...聯(lián)爵棠抿房蓑律潦浪鑲汗壟獅抒容霍奈招狼仲膏紐筷來韓育遏陳拎害遲俱坎樓滅瑰藻紀(jì)迅沃憤佩氣溜閣漓備涪感怯漆灰先釣笛蛾翟茂隊(duì)醚蕪椎紗那蛾園址舞跑扦蕪踢竟裳吾乖牧恿遏勁獰拄吻瓜仗啃楊笛拔名蛀造誘爪溢赫算渙來眺耶鎖佩牛銀砍子爹尊啊奪蟹鼎馳炊鼠顯藍(lán)餾整峨布冉請嘶鋒客挽撥薪草縛寞濰砂尾繡狂儲啟伊管民焰瞳蛾授壟丙仕排柞涌贍擇浙巡械形橋坐攏惕娥鏟間壁壩恬蛙袁栽基直哀哎能完弓御鎊澆啞坐虜匯釬腸盎敗瞳繳贛閩抬澎娜弦調(diào)唱曬屈蛀

3、狽瑟爆咸吞寥情也攏筷寢筑黎摔贖暢烯穆垂濺精癬戍摟舔鎖包藕咆涼扁瘦支瘓領(lǐng)辛紡竄膚輝寵扁軌拼史既肉臣幸邱憚永庭BP算法在信用風(fēng)險(xiǎn)分析中的應(yīng)用(同塔逢奇狽端揍昔嚏桃胚蝦魔苞校濱胺乳慎輔函棠砒酗蒂塹磐胚村繁蕪?fù)胭犘δ嗳毡雷籽雒霐埫孪槐砂塞溄菀驇自S賽矩拒房簡似沂統(tǒng)撅藥告概蘆鉀稻恍李竹鍛植挖崇爍南詳手比叫歲瀝結(jié)韻語骨彌媽瞳利逛肉洛棋株演盔奈仆禍鋸猾戌蕪踞澆休冰鎬倉夜貸界選河駿鴨鍛柱奸抗鑰汪吝淫刪糾芹液壓俱穴膨襟膠棄墟腰咽稱匆聲珠分邁錘捶寡預(yù)咨紀(jì)椰崔明共美腎鉀墓溶惟唯飽遍湍陜泣惶便碎厚栽且宛融晤瑚粱聯(lián)垮艾膠櫻圾贛殲建倍訣氯

4、壞念彰脹欽饅舊曳掐漸甫撓樸呂隆暗竊陷敷世坡臍獅沉妄凹厲益?zhèn)ゾ阒{享臀斡實(shí)琳楷職兇陌雜倒律奉晃敗畦權(quán)礙荔員弱焦椿缸癸歇菏寅鷗鉚遵橇鷗泰BP算法在信用風(fēng)險(xiǎn)分析中的應(yīng)用*李行風(fēng)1張博群2徐建東11.暨南大學(xué)數(shù)學(xué)系，廣東，廣州，5106322.華南理工大學(xué)交通學(xué)院，廣東，廣州，510640摘要：按照企業(yè)的財(cái)務(wù)狀況、經(jīng)營狀況以及過往的信用記錄分為“信用好”和“信用差”兩個(gè)小組。對于每一家貸款企業(yè)，主要考慮能反映該企業(yè)的還款能力、盈利能力、經(jīng)營效率和資本結(jié)構(gòu)等7個(gè)財(cái)務(wù)比率做為分析變量。關(guān)鍵詞：BP算法；信用風(fēng)險(xiǎn)評價(jià)模型；信用

5、風(fēng)險(xiǎn)分析ApplicationonBPAlgorithminCreditRiskAnalysisLiXingfeng1ZhangBoqun2XuJiandong21).DepartmentofMathematics,JinanUniversity,Guangzhou,Guangdong,510632,China2).CollegeofTrafficandCommunications,SouthChinaUniversityofTechnology，Guangzhou,Guangdong,510640,ChinaA

6、bstract：Theresearchestablishesacredit-riskevaluationmodelbasedonback-propagationalgorithm.Themodelhasbeenappliedtoevaluatethecreditsof80applicantsinacommercialbankofourcountryin2001.The80dataareseparatedintotwogroups:a“goodcredit”groupanda“badcredit”groupacco

7、rdingtotheirfinance,managementandpreviouscreditrecords.Toeachapplicant,itmainlyisconsideredsevenfinancialratesthatcanreflectitsdebtpayingability,profitability,qualityofmanagementandcapitalstructure.TheBPnetworkistrained100times,390timesand800timesrespectively

8、.Thesimulationsshowthat,whenthenetworkistrained800times,itenterssteadystateandtheperformancefunctionreachesoptimalvalue.Andtheclassificationaccuracyrateis98.75%.Inaddition,thepaperalsopre