Predicting repayment of borrows in peer-to-peer social lending with deep dense convolutional network

Ji Yoon Kim, Sung Bae Cho

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

In peer-to-peer lending, it is important to predict the repayment of the borrower to reduce the lender's financial loss. However, it is difficult to design a powerful feature extractor for predicting the repayment as user and transaction data continue to increase. Convolutional neural networks automatically extract useful features from big data, but they use only high-level features; hence, it is difficult to capture a variety of representations. In this study, we propose a deep dense convolutional network for repayment prediction in social lending, which maintains the borrower's semantic information and obtains a good representation by automatically extracting important low- and high-level features simultaneously. We predict the repayment of the borrower by learning discriminative features depending on the loan status. Experimental results on the Lending Club dataset show that our model is more effective than other methods. A fivefold cross-validation is performed to run the experiments.

Original languageEnglish
Article numbere12403
JournalExpert Systems
Volume36
Issue number4
DOIs
Publication statusPublished - 2019 Jan 1

Fingerprint

Peer to Peer
Semantics
Neural networks
Predict
Extractor
Cross-validation
Transactions
Continue
Experiments
Neural Networks
Prediction
Experimental Results
Experiment
Big data
Model
Learning
Design

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Theoretical Computer Science
  • Computational Theory and Mathematics
  • Artificial Intelligence

Cite this

@article{f2e1da73668e49d3af330741c64d1ade,
title = "Predicting repayment of borrows in peer-to-peer social lending with deep dense convolutional network",
abstract = "In peer-to-peer lending, it is important to predict the repayment of the borrower to reduce the lender's financial loss. However, it is difficult to design a powerful feature extractor for predicting the repayment as user and transaction data continue to increase. Convolutional neural networks automatically extract useful features from big data, but they use only high-level features; hence, it is difficult to capture a variety of representations. In this study, we propose a deep dense convolutional network for repayment prediction in social lending, which maintains the borrower's semantic information and obtains a good representation by automatically extracting important low- and high-level features simultaneously. We predict the repayment of the borrower by learning discriminative features depending on the loan status. Experimental results on the Lending Club dataset show that our model is more effective than other methods. A fivefold cross-validation is performed to run the experiments.",
author = "Kim, {Ji Yoon} and Cho, {Sung Bae}",
year = "2019",
month = "1",
day = "1",
doi = "10.1111/exsy.12403",
language = "English",
volume = "36",
journal = "Expert Systems",
issn = "0266-4720",
publisher = "Wiley-Blackwell",
number = "4",

}

Predicting repayment of borrows in peer-to-peer social lending with deep dense convolutional network. / Kim, Ji Yoon; Cho, Sung Bae.

In: Expert Systems, Vol. 36, No. 4, e12403, 01.01.2019.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Predicting repayment of borrows in peer-to-peer social lending with deep dense convolutional network

AU - Kim, Ji Yoon

AU - Cho, Sung Bae

PY - 2019/1/1

Y1 - 2019/1/1

N2 - In peer-to-peer lending, it is important to predict the repayment of the borrower to reduce the lender's financial loss. However, it is difficult to design a powerful feature extractor for predicting the repayment as user and transaction data continue to increase. Convolutional neural networks automatically extract useful features from big data, but they use only high-level features; hence, it is difficult to capture a variety of representations. In this study, we propose a deep dense convolutional network for repayment prediction in social lending, which maintains the borrower's semantic information and obtains a good representation by automatically extracting important low- and high-level features simultaneously. We predict the repayment of the borrower by learning discriminative features depending on the loan status. Experimental results on the Lending Club dataset show that our model is more effective than other methods. A fivefold cross-validation is performed to run the experiments.

AB - In peer-to-peer lending, it is important to predict the repayment of the borrower to reduce the lender's financial loss. However, it is difficult to design a powerful feature extractor for predicting the repayment as user and transaction data continue to increase. Convolutional neural networks automatically extract useful features from big data, but they use only high-level features; hence, it is difficult to capture a variety of representations. In this study, we propose a deep dense convolutional network for repayment prediction in social lending, which maintains the borrower's semantic information and obtains a good representation by automatically extracting important low- and high-level features simultaneously. We predict the repayment of the borrower by learning discriminative features depending on the loan status. Experimental results on the Lending Club dataset show that our model is more effective than other methods. A fivefold cross-validation is performed to run the experiments.

UR - http://www.scopus.com/inward/record.url?scp=85063892165&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85063892165&partnerID=8YFLogxK

U2 - 10.1111/exsy.12403

DO - 10.1111/exsy.12403

M3 - Article

AN - SCOPUS:85063892165

VL - 36

JO - Expert Systems

JF - Expert Systems

SN - 0266-4720

IS - 4

M1 - e12403

ER -