復雜擔保網(wǎng)絡的結(jié)構(gòu)性質(zhì)和風險分析
發(fā)布時間：2018-11-04 07:51
【摘要】：隨著經(jīng)濟一體化的高速發(fā)展,擔保作為一種很重要的籌集資金方式,在資金融通中被廣泛使用,但是由于其復雜且不透明特性,已成為加劇信貸風險暴露的重要原因,引起社會各界普遍關(guān)注。由于傳統(tǒng)分析方法對處理大規(guī)模擔保網(wǎng)絡問題的局限性,復雜網(wǎng)絡等大數(shù)據(jù)技術(shù)則為擔保圈風險分析提供了契機。本文利用某金融機構(gòu)提供的擔保數(shù)據(jù),基于復雜網(wǎng)絡理論分析擔保網(wǎng)絡的拓撲結(jié)構(gòu),重點識別擔保圈的社團結(jié)構(gòu),以及揭示擔保風險的傳播機制,為有針對性的進行風險防范,降低風險管理成本提供幫助。具體的研究內(nèi)容可以歸納為以下幾方面:(1)提出一種快速的社團挖掘算法Fast-SCAN(FSCAN)。SCAN算法的時間復雜度是O(4|E|2/|V|),針對該算法在大規(guī)模網(wǎng)絡社團劃分中時間開銷較高的問題,本文改進后的時間復雜度是O(2|E|2/|V|-2K|E|/|V|),能夠高效發(fā)現(xiàn)大規(guī)模網(wǎng)絡中的社區(qū)結(jié)構(gòu)、集群點以及噪音點。(2)本文引入并查集思想,提出基于路徑壓縮的譜系劃分算法,可以發(fā)現(xiàn)擔保網(wǎng)絡中連通的擔保圈。結(jié)合圖論和復雜網(wǎng)絡理論對其結(jié)構(gòu)進行分析,同時使用FSCAN算法識別核心擔保圈的社團結(jié)構(gòu)。(3)針對待查信用主體在擔保網(wǎng)絡中隱藏的間接擔保關(guān)系,提出一種模擬鄰接表的數(shù)據(jù)結(jié)構(gòu),同時融合圖論的遍歷思想,提出待查信用主體K代關(guān)聯(lián)關(guān)系的路徑發(fā)現(xiàn)算法,發(fā)現(xiàn)待查信用主體之間隱藏的間接擔保關(guān)系,以防范擔保圈風險的爆發(fā)。(4)通過對擔保圈的風險傳播特性進行分析,改進災害蔓延動力學模型,提出更適用的風險傳播模型。仿真?zhèn)鞑ピ磳︼L險傳播的影響,對待查信用主體進行信用評級,并仿真關(guān)鍵企業(yè)在核心擔保群中風險傳播的路徑。最后,提出相應監(jiān)管策略,為防范因擔保圈引發(fā)的金融危機提供理論參考。
[Abstract]:With the rapid development of economic integration, guarantee, as a very important way to raise funds, is widely used in financing, but because of its complex and opaque characteristics, it has become an important reason for increasing exposure to credit risk. Cause the general concern of all walks of life of the society. Because of the limitation of traditional analysis method to deal with the problem of large-scale guarantee network, big data technology such as complex network provides an opportunity for risk analysis of guarantee circle. Based on the theory of complex network, this paper analyzes the topological structure of the guarantee network by using the guarantee data provided by a financial institution, focuses on identifying the community structure of the guarantee circle, and reveals the spreading mechanism of the guarantee risk in order to guard against the risk. Help reduce risk management costs. The specific research contents can be summarized as follows: (1) the time complexity of a fast community mining algorithm Fast-SCAN (FSCAN). SCAN is O (4E _ 2 / V). In order to solve the problem of high time cost in mass network community partitioning, the time complexity of the improved algorithm is O (2E 2 / C-V 2K E / V), which can efficiently discover the community structure in large scale network. Cluster points and noise points. (2) in this paper, we introduce and search the idea of set, and propose a pedigree partition algorithm based on path compression, which can find the connected guarantee circle in the guaranteed network. Combined with graph theory and complex network theory, this paper analyzes its structure, and uses FSCAN algorithm to identify the community structure of the core guarantee circle. (3) aiming at the indirect guarantee relationship hidden by the credit subject in the guarantee network, In this paper, we propose a data structure that simulates the adjacency table, and combine the traversal idea of graph theory, and propose a path discovery algorithm for K-generation association relation of the credit subject to be checked, and find the hidden indirect guarantee relationship between the credit subject under check. (4) by analyzing the risk propagation characteristics of the guarantee circle, the dynamic model of disaster spread is improved, and a more suitable risk propagation model is put forward. This paper simulates the influence of the source of communication on the risk propagation, evaluates the credit rating of the credit subject, and simulates the path of risk transmission in the core guarantee group of key enterprises. Finally, the corresponding regulatory strategies are put forward to provide a theoretical reference for preventing the financial crisis caused by the guarantee circle.
【學位授予單位】：大連海事大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：O157.5
[Abstract]:With the rapid development of economic integration, guarantee, as a very important way to raise funds, is widely used in financing, but because of its complex and opaque characteristics, it has become an important reason for increasing exposure to credit risk. Cause the general concern of all walks of life of the society. Because of the limitation of traditional analysis method to deal with the problem of large-scale guarantee network, big data technology such as complex network provides an opportunity for risk analysis of guarantee circle. Based on the theory of complex network, this paper analyzes the topological structure of the guarantee network by using the guarantee data provided by a financial institution, focuses on identifying the community structure of the guarantee circle, and reveals the spreading mechanism of the guarantee risk in order to guard against the risk. Help reduce risk management costs. The specific research contents can be summarized as follows: (1) the time complexity of a fast community mining algorithm Fast-SCAN (FSCAN). SCAN is O (4E _ 2 / V). In order to solve the problem of high time cost in mass network community partitioning, the time complexity of the improved algorithm is O (2E 2 / C-V 2K E / V), which can efficiently discover the community structure in large scale network. Cluster points and noise points. (2) in this paper, we introduce and search the idea of set, and propose a pedigree partition algorithm based on path compression, which can find the connected guarantee circle in the guaranteed network. Combined with graph theory and complex network theory, this paper analyzes its structure, and uses FSCAN algorithm to identify the community structure of the core guarantee circle. (3) aiming at the indirect guarantee relationship hidden by the credit subject in the guarantee network, In this paper, we propose a data structure that simulates the adjacency table, and combine the traversal idea of graph theory, and propose a path discovery algorithm for K-generation association relation of the credit subject to be checked, and find the hidden indirect guarantee relationship between the credit subject under check. (4) by analyzing the risk propagation characteristics of the guarantee circle, the dynamic model of disaster spread is improved, and a more suitable risk propagation model is put forward. This paper simulates the influence of the source of communication on the risk propagation, evaluates the credit rating of the credit subject, and simulates the path of risk transmission in the core guarantee group of key enterprises. Finally, the corresponding regulatory strategies are put forward to provide a theoretical reference for preventing the financial crisis caused by the guarantee circle.
【學位授予單位】：大連海事大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：O157.5
【相似文獻】
相關(guān)會議論文 前1條
1 郭小平;;“風險傳播”研究的范式轉(zhuǎn)換[A];2006中國傳播學論壇論文集（Ⅰ）[C];2006年
相關(guān)重要報紙文章 前1條
1 劉濤;公民行動與風險傳播[N];中國社會科學院報;2009年
相關(guān)博士學位論文 前1條
1 朱閔敏;肺結(jié)核風險傳播干預與效果評估[D];華中科技大學;2013年
相關(guān)碩士學位論文 前10條
1 程宇婕;風險傳播理論研究[D];山西大學;2010年
2 楊佳彤;媒體在科技風險傳播中的效應研究[D];西北農(nóng)林科技大學;2015年
3 王佳樂;茂名PX環(huán)境沖突事件：環(huán)境風險傳播中政府、公眾與媒體的角色互動[D];陜西師范大學;2015年
4 韓冰潔;基于WBS的復雜產(chǎn)品制造過程風險傳播與管理研究[D];江蘇科技大學;2015年
5 裴思e，
本文編號：2309187
本文鏈接：http://sikaile.net/shoufeilunwen/benkebiyelunwen/2309187.html
最近更新
教材專著
