Cost-sensitive boosted tree for loan evaluation in peer-to-peer lending

Xia, Yufei; Liu, Chuanzhe; Liu, Nana

doi:10.1016/j.elerap.2017.06.004

Cited by 150 publications

(113 citation statements)

References 53 publications

Supporting

Mentioning

101

Contrasting

Order By: Relevance

“…Nevertheless, the parameters of XGBoost model are manually and empirically tuned in this study. Therefore, better performance can be obtained by optimizing the tuning parameters using Bayesian [27], genetic algorithms [45], and any other methods in future works.…”

Section: Discussionmentioning

confidence: 99%

“…In this research, eXtreme Gradient Boosting (XGBoost) is selected as the estimation method of statistical line loss of distribution feeders, since it is extensively used by scientific researchers and engineers, and has remarkable performance in many fields, such as energy and remote sensing [15][16][17][18], information technology and software engineering [19][20][21][22], biological and medical engineering [23][24][25], economy, and finance [26,27]. Except for the precise and robust model that is established in XGBoost [28], it is flexible to rewrite the objective function of XGBoost, which makes it possible to estimate statistical line loss with reference to the theoretical value.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Novel Method of Statistical Line Loss Estimation for Distribution Feeders Based on Feeder Cluster and Modified XGBoost

2017

View full text Add to dashboard Cite

Abstract:The estimation of losses of distribution feeders plays a crucial guiding role for the planning, design, and operation of a distribution system. This paper proposes a novel estimation method of statistical line loss of distribution feeders using the feeder cluster technique and modified eXtreme Gradient Boosting (XGBoost) algorithm that is based on the characteristic data of feeders that are collected in the smart power distribution and utilization system. In order to enhance the applicability and accuracy of the estimation model, k-medoids algorithm with weighting distance for clustering distribution feeders is proposed. Meanwhile, a variable selection method for clustering distribution feeders is discussed, considering the correlation and validity of variables. This paper next modifies the XGBoost algorithm by adding a penalty function in consideration of the effect of the theoretical value to the loss function for the estimation of statistical line loss of distribution feeders. The validity of the proposed methodology is verified by 762 distribution feeders in the Shanghai distribution system. The results show that the XGBoost method has higher accuracy than decision tree, neural network, and random forests by comparison of Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and Absolute Percentage Error (APE) indexes. In particular, the theoretical value can significantly improve the reasonability of estimated results.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Novel Method of Statistical Line Loss Estimation for Distribution Feeders Based on Feeder Cluster and Modified XGBoost

2017

View full text Add to dashboard Cite

show abstract

“…Furthermore, two other groups, which can be designed based on either the data level or algorithm level methods, or both, are also commonly applied in existing studies. They are cost-sensitive learning techniques (Xia et al 2017;Chao and Peng 2018) and ensemble methods (Galar et al 2012;Galar et al 2013;He et al 2018). Generally speaking, algorithm level methods are more dependent on the problems, and data level-based resampling techniques are independent of the underlying classifier.…”

Section: Resampling Techniques For Class Imbalancementioning

confidence: 99%

Malicious web domain identification using online credibility and performance data by considering the class imbalance issue

Chiong

Pranata

et al. 2019

IMDS

View full text Add to dashboard Cite

Purpose -Malicious web domain identification is of significant importance to the security protection of Internet users. With online credibility and performance data, this paper aims to investigate the use of machine learning techniques for malicious web domain identification by considering the class imbalance issue (i.e., there are more benign web domains than malicious ones). Design/methodology/approach -We propose an integrated resampling approach to handle class imbalance by combining the Synthetic Minority Over-sampling TEchnique (SMOTE) and Particle Swarm Optimisation (PSO), a population-based meta-heuristic algorithm. We use the SMOTE for over-sampling and PSO for under-sampling. Findings -By applying eight well-known machine learning classifiers, the proposed integrated resampling approach is comprehensively examined using several imbalanced web domain datasets with different imbalance ratios. Compared to five other well-known resampling approaches, experimental results confirm that the proposed approach is highly effective. Practical implications -This study not only inspires the practical use of online credibility and performance data for identifying malicious web domains, but also provides an effective resampling approach for handling the class imbalance issue in the area of malicious web domain identification. Originality/value -Online credibility and performance data is applied to build malicious web domain identification models using machine learning techniques. An integrated resampling approach is proposed to address the class imbalance issue. The performance of the proposed approach is confirmed based on real-world datasets with different imbalance ratios.

show abstract

“…Peer-to-peer (P2P) lending, also known as social lending, has become more popular in recent years because it provides an online trading platform for lending money without the interference of traditional intermediaries, such as banks. Social lending companies reduce the cost of finance by connecting the lender and the borrower directly [1]. Despite the sophisticated mechanisms these platforms provide to evaluate credit risk, social lenders still face the risks associated with unsecured loans [2].…”

Section: Introductionmentioning

confidence: 99%

“…Moreover, from a profitability standpoint, recognizing default loans is not only critical for lenders but also for the sustainability of the P2P lending market. Therefore, correctly identifying credit risk to support decision making about the eligibility of a particular borrower has emerged as a critical problem for P2P lending platforms [1,3].…”

Section: Introductionmentioning

confidence: 99%

Handling Uncertainty in Social Lending Credit Risk Prediction with a Choquet Fuzzy Integral Model

Namvar

Naderpour

2018

2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)

View full text Add to dashboard Cite

As one of the main business models in the financial technology field, peer-to-peer (P2P) lending has disrupted traditional financial services by providing an online platform for lending money that has remarkably reduced financial costs. However, the inherent uncertainty in P2P loans can result in huge financial losses for P2P platforms. Therefore, accurate risk prediction is critical to the success of P2P lending platforms. Indeed, even a small improvement in credit risk prediction would be of benefit to P2P lending platforms. This paper proposes an innovative credit risk prediction framework that fuses base classifiers based on a Choquet fuzzy integral. Choquet integral fusion improves creditworthiness evaluations by synthesizing the prediction results of multiple classifiers and finding the largest consistency between outcomes among conflicting and consistent results. The proposed model was validated through experimental analysis on a realworld dataset from a well-known P2P lending marketplace. The empirical results indicate that the combination of multiple classifiers based on fuzzy Choquet integrals outperforms the best base classifiers used in credit risk prediction to date. In addition, the proposed methodology is superior to some conventional combination techniques. Keywords-Choquet fuzzy integral; fuzzy measure; credit risk prediction; peer-to-peer lendingI.

show abstract

Cost-sensitive boosted tree for loan evaluation in peer-to-peer lending

Cited by 150 publications

References 53 publications

A Novel Method of Statistical Line Loss Estimation for Distribution Feeders Based on Feeder Cluster and Modified XGBoost

A Novel Method of Statistical Line Loss Estimation for Distribution Feeders Based on Feeder Cluster and Modified XGBoost

Malicious web domain identification using online credibility and performance data by considering the class imbalance issue

Handling Uncertainty in Social Lending Credit Risk Prediction with a Choquet Fuzzy Integral Model

Contact Info

Product

Resources

About