Support Vector Machine (SVM) as Financial Distress Model Prediction in Property and Real Estate Companies

Abstract

Financial distress prediction is an interesting topic to be studied because of its significant impact on various stakeholders. Various methods have been developed to predict the company's financial distress. Among the famous models, the Support Vector Machine (SVM) is claimed to be the most successful model in prediction and classification. SVM is a machine learning method that works on the principle of Structural Risk Minimization (SRM) with the aim of finding the best hyperplane that separates two classes in the input space by maximizing the hyperplane margin and obtaining the best support vector. This study applies the SVM model in predicting the financial distress of property and real estate companies listed on the Indonesia Stock Exchange. There were 18 variables of financial ratios used in this study. By Using Principal Component Analysis (PCA) in feature selections there are five variables selected in this study, namely Return on Assets, Return on Equity, Net Profit Margin, Earning Per Share, and Operating Profit Margin. The SVM model is formed by dividing the training and testing data with 10-fold cross-validation and using three kernels: linear kernel, polynomial, and Radial Basis Function (RBF). The best SVM model formed is the SVM model with RBF kernel type with parameters sigma = 1 and C = 1.0 which can predict financial distress with an accuracy value of 82.99% and an error rate of 17.01%.

A Hybrid Network Analysis and Machine Learning Model for Enhanced Financial Distress Prediction

“…Machine learning algorithms have become extensively utilized in predicting company financial difficulties in recent years [21]. Several previous studies have applied the support vector machine prediction model in the field of financial distress [22]. In addition, the rapid advancement of computers and software has given rise to other techniques such as data mining, machine learning, deep learning, and artificial intelligence [23].…”

“…Several studies have highlighted the superiority of machine learningbased models over traditional methods in predicting corporate financial distress [11]. These machine learning techniques include support vector machines, deep learning models, hybrid machine learning technologies, genetic algorithms, and neural network models [21], [22], [20], [27], [28]. [29] compared traditional methods such as logistic regression with machine learning models like random forest and neural networks to identify the model with the highest predictive accuracy of financial distress.…”

“…Hybrid machine learning techniques have also been explored for financial distress prediction, aiming to establish effective prediction models. focused on developing an effective financial distress prediction model using hybrid machine learning techniques [22]. In addition, [22] introduced a novel framework for a financial early warning system that enhances its effectiveness by integrating the unconstrained distributed lag model with commonly employed financial distress prediction models, such as the logistic model and SVM.…”

Kmeans-SMOTE Integration for Handling Imbalance Data in Classifying Financial Distress Companies using SVM and Naïve Bayes

“…Several studies have been conducted on the classification of financial distress using SVM and Naive Bayes algorithms. The study [4] utilized the SVM algorithm without hyperparameter tuning for the classification of financial distress. The results of the research showed an accuracy of 81.06%, an error rate of 18.94%, a precision of 89.09%, and a recall of 59.04%.…”