Comparative Analysis of RF, SVR with Gaussian Kernel and LSTM for Predicting Loan Defaults

This investigation elucidates the paramount endeavour of predicting loan defaults, which is imperative for the efficacious management of financial risk and the overall stability of financial institutions. Conventional statistical methodologies frequently encounter challenges in effectively capturing...

Full description

Saved in:
Bibliographic Details
Main Authors: Konstantinos Kofidis, Cătălina Lucia Cocianu
Format: Article
Language:English
Published: Institutul de Studii Financiare 2024-11-01
Series:Revista de Studii Financiare
Subjects:
Online Access:https://revista.isfin.ro/wp-content/uploads/2024/11/6.-Konstantinos-K.-et..pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849429525227634688
author Konstantinos Kofidis
Cătălina Lucia Cocianu
author_facet Konstantinos Kofidis
Cătălina Lucia Cocianu
author_sort Konstantinos Kofidis
collection DOAJ
description This investigation elucidates the paramount endeavour of predicting loan defaults, which is imperative for the efficacious management of financial risk and the overall stability of financial institutions. Conventional statistical methodologies frequently encounter challenges in effectively capturing the nonlinear and sequential dynamics inherent in financial data, thereby necessitating the examination of more sophisticated machine learning methodologies. This research reports an experimental-based comparative evaluation of three ML and DL models—Long Short-Term Memory (LSTM) networks, Random Forest (RF), and Support Vector Regression (SVR)—to assess their efficacy in forecasting loan defaults. The models are evaluated using metrics such as Mean Squared Error (MSE), F1 score, and Accuracy, and their proficiency in addressing imbalanced datasets and elucidating intricate data relationships is highlighted. The results indicate that while the Random Forest model surpasses its counterparts in terms of accuracy and MSE, the LSTM model exhibits considerable potential in managing imbalanced data, as evidenced by its stable F1 score. Although SVR reveals competitive precision, it exhibits deficiencies in addressing class imbalance. The ANOVA analyses substantiate that the disparities in model performance are statistically significant. The research acknowledges that both the LSTM and SVR models remain in the developmental stages, with ongoing initiatives aimed at refining these models through hyperparameter optimization and advanced architectural frameworks to enhance their predictive efficacy in practical applications.
format Article
id doaj-art-2853703fe5b84e8d980a1d0b3e457ef2
institution Kabale University
issn 2537-3714
2559-1347
language English
publishDate 2024-11-01
publisher Institutul de Studii Financiare
record_format Article
series Revista de Studii Financiare
spelling doaj-art-2853703fe5b84e8d980a1d0b3e457ef22025-08-20T03:28:21ZengInstitutul de Studii FinanciareRevista de Studii Financiare2537-37142559-13472024-11-019179110610.55654/JFS.2024.9.17.06Comparative Analysis of RF, SVR with Gaussian Kernel and LSTM for Predicting Loan DefaultsKonstantinos Kofidis0Cătălina Lucia Cocianu1Bucharest University of Economic Studies, Bucharest, RomaniaBucharest University of Economic Studies, Bucharest, RomaniaThis investigation elucidates the paramount endeavour of predicting loan defaults, which is imperative for the efficacious management of financial risk and the overall stability of financial institutions. Conventional statistical methodologies frequently encounter challenges in effectively capturing the nonlinear and sequential dynamics inherent in financial data, thereby necessitating the examination of more sophisticated machine learning methodologies. This research reports an experimental-based comparative evaluation of three ML and DL models—Long Short-Term Memory (LSTM) networks, Random Forest (RF), and Support Vector Regression (SVR)—to assess their efficacy in forecasting loan defaults. The models are evaluated using metrics such as Mean Squared Error (MSE), F1 score, and Accuracy, and their proficiency in addressing imbalanced datasets and elucidating intricate data relationships is highlighted. The results indicate that while the Random Forest model surpasses its counterparts in terms of accuracy and MSE, the LSTM model exhibits considerable potential in managing imbalanced data, as evidenced by its stable F1 score. Although SVR reveals competitive precision, it exhibits deficiencies in addressing class imbalance. The ANOVA analyses substantiate that the disparities in model performance are statistically significant. The research acknowledges that both the LSTM and SVR models remain in the developmental stages, with ongoing initiatives aimed at refining these models through hyperparameter optimization and advanced architectural frameworks to enhance their predictive efficacy in practical applications.https://revista.isfin.ro/wp-content/uploads/2024/11/6.-Konstantinos-K.-et..pdfloan default predictionfinancial risk managementlong short-term memory (lstm)support vector regression (svr)mean squared errorf1 score
spellingShingle Konstantinos Kofidis
Cătălina Lucia Cocianu
Comparative Analysis of RF, SVR with Gaussian Kernel and LSTM for Predicting Loan Defaults
Revista de Studii Financiare
loan default prediction
financial risk management
long short-term memory (lstm)
support vector regression (svr)
mean squared error
f1 score
title Comparative Analysis of RF, SVR with Gaussian Kernel and LSTM for Predicting Loan Defaults
title_full Comparative Analysis of RF, SVR with Gaussian Kernel and LSTM for Predicting Loan Defaults
title_fullStr Comparative Analysis of RF, SVR with Gaussian Kernel and LSTM for Predicting Loan Defaults
title_full_unstemmed Comparative Analysis of RF, SVR with Gaussian Kernel and LSTM for Predicting Loan Defaults
title_short Comparative Analysis of RF, SVR with Gaussian Kernel and LSTM for Predicting Loan Defaults
title_sort comparative analysis of rf svr with gaussian kernel and lstm for predicting loan defaults
topic loan default prediction
financial risk management
long short-term memory (lstm)
support vector regression (svr)
mean squared error
f1 score
url https://revista.isfin.ro/wp-content/uploads/2024/11/6.-Konstantinos-K.-et..pdf
work_keys_str_mv AT konstantinoskofidis comparativeanalysisofrfsvrwithgaussiankernelandlstmforpredictingloandefaults
AT catalinaluciacocianu comparativeanalysisofrfsvrwithgaussiankernelandlstmforpredictingloandefaults