Methods of a decision support system for investment in peer to peer lending network

Abstract

At present, FinTech technology has started to create financial innovations for Thailand, making transactions via online channels easier and faster. One financial innovation getting attention is the P2P lending platform that will replace traditional borrowing. All these processes are carried out on all online platforms and act as an online credit market where credit contract transactions between borrowers and investors can meet directly. This new economic lending model presents unique challenges in making effective investment decisions. The researchers note a significant problem with the risk of default in P2P lending platforms. The researcher sought to study a sound decision support system for risk assessment based on the above issues. This research aims to examine the methods of decision support systems for investing in P2P lending networks. This research is quantitative. The researchers selected data from Prosper Marketplace, Inc. for data analysis. Additionally, the researchers chose to use the Bipartite Graph Theory to explain the behavior of P2P borrowing and the weighted average equation to calculate the level of investor confidence. To estimate the probability of payment of all new loans is then applied to the next step of analyzing the results for platform performance. A study was conducted to analyze the decision support system model for repayment confidence. For this paper, the researchers wanted to increase the trend of higher payback confidence as a strategy to support the P2P lending platform. Four decision models were compared: Linear regression, Logistic regression, Bayesian regression, and Support Vector Machines. The results showed that the decision model of the Support Vector Machine model tended to have a higher confidence rate than other models. In terms of strategy, it can attract investors to make investment decisions through P2P lending platforms. The remaining three models were close to the actual average in terms of efficiency and stability. These sentiment trends can help investors make the best investment decisions to support them before investment decisions.IMPLICATION OF THE THEMATIC PAPER. The main implication of this study is to explain how the Bipartite Graph Theory can well explain P2P lending algorithms for matching between investors and borrowers. It can be used as an analytical tool to guide investors in making effective investment decisions for both investors and borrowers. The study then found a model that best supports decision-making to explain strategic repayment. Other future borrowing approaches may be proposed using the Blockchain platform for credit scoring decisions in conjunction with Bipartite Graph Theory. This allows for further improvements applied to the P2P lending platforms to be reliably verified. In the real world, it also increases the monitoring of P2P lending algorithms by considering the Multi-Objective Evolutionary Algorithm (MOEA), as the efficacy of MOEAs has been confirmed in several issues.

ในปัจจุบันเทคโนโลยี ฟินเทค ได้เริ่มเข้ามาสร้างนวัตกรรมทางการเงินให้กับประเทศไทย ทำให้การทำธุรกรรมผ่านทางช่องทางออนไลน์นั้นง่ายและสะดวกเร็วขึ้น ซึ่งหนึ่งในนวัตกรรมทางการเงินที่กำลังได้รับความสนใจอยู่ในขณะนี้ คือ แพลตฟอร์มเพียร์ทูเพียร์เลนดิง ที่จะเข้ามาแทนการกู้ยืมเงินแบบเดิม กระบวนการทั้งหมดเหล่านี้จะ