Analisis Perbandingan Klasifikasi Financial Distress pada Perusahaan Sektor Perbankan dengan Artificial Neural Network dan Support Vector Machine

Apriliani, Dina Silmi (2024) Analisis Perbandingan Klasifikasi Financial Distress pada Perusahaan Sektor Perbankan dengan Artificial Neural Network dan Support Vector Machine. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5006201017_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (3MB) | Request a copy

Abstract

Pengelolaan keuangan semakin mendesak seiring dengan meningkatnya inflasi dan kebutuhan hidup manusia. Investasi menjadi solusi populer, namun juga membawa risiko penipuan bagi investor pemula. Pemahaman yang baik tentang kondisi perusahaan menjadi penting, khususnya dalam sektor perbankan yang memiliki dampak signifikan terhadap stabilitas keuangan. Pandemi COVID-19 pada tahun 2020 menimbulkan tantangan besar dengan memengaruhi kinerja perusahaan perbankan dan menimbulkan potensi risiko Financial Distress. Financial Distress adalah kondisi di mana perusahaan tidak memiliki cukup kas operasional untuk melanjutkan bisnisnya dan gagal memenuhi kewajiban debitur. Dalam menghadapi tantangan ini, teknik klasifikasi data menjadi penting, dengan metode klasifikasi seperti Artificial Neural Network (ANN) dan Support Vector Machine (SVM) terbukti efektif dalam analisis klasifikasi. Penelitian ini bertujuan untuk menganalisis klasifikasi perusahaan sektor perbankan di BEI menggunakan metode ANN dan SVM dengan data laporan keuangan sektor perbankan dari tahun 2018 hingga 2022. Klasifikasi awal dalam mengidentifikasi perusahaan perbankan yang mengalami Financial Distress dilakukan dengan menggunakan Altman Modifikasi. Berdasarkan analisis yang telah dilakukan model ANN Multilayer Perceptron (MLP) dengan Backpropagation yang terbaik adalah 15 layer input, 6 hidden neuron, dan 1 hidden layer, mencapai F1-Score 96% untuk kelas 0 dan 93% untuk kelas 1 pada data training. Sedangkan pada data testing, 94% untuk kelas 0 dan 80% untuk kelas 1. Sedangkan klasifikasi menggunakan Support Vector Machine (SVM) dibagi menjadi 3 yaitu kernel Radial Basis Function (RBF), Polynomial, dan Sigmoid. Model Support Vector Machine (SVM) terbaik adalah Support Vector Machine dengan kernel Sigmoid dengan nilai parameter C=100 dan γ=0,01 memiliki nilai F1-Score 95% untuk kelas 0 dan 92% untuk kelas 1 pada data training. Sedangkan pada data testing 89% untuk kelas 0 dan 71% untuk kelas 1. ANN menunjukkan konsistensi yang baik antara F1-Score pada data training dan testing dengan perbedaan kecil. Meskipun SVM dengan kernel Sigmoid menunjukkan nilai F1-Score yang baik, namun ANN unggul dalam data testing.

Kata kunci: Altman Modifikasi, Artificial Neural Network, F1-Score, Financial Distress, Support Vector Machine.

============================================================

Financial management is increasingly urgent in line with increasing inflation and human living needs. Investing is a popular solution, but it also carries the risk of fraud for novice investors. A good understanding of the company's condition is important, especially in the banking sector which has a significant impact on financial stability. The COVID-19 pandemic in 2020 posed a major challenge by affecting the performance of banking companies and posing a potential risk of Financial Distress. Financial Distress is a condition in which a company does not have enough operating cash to continue its business and fails to meet debtor obligations. In the face of these challenges, data classification techniques are important, with classification methods such as Artificial Neural Network (ANN) and Support Vector Machine (SVM) proving effective in classification analysis. This study aims to analyze the classification of banking sector companies on the IDX using the ANN and SVM methods with financial statement data of the banking sector from 2018 to 2022. The initial classification in identifying banking companies experiencing Financial Distress was carried out using Altman Modification. Based on the analysis that has been carried out, the ANN Multilayer Perceptron (MLP) model with the best backpropagation is 15 input layers, 6 hidden neurons, and 1 hidden layer, achieving an F1-Score of 96% for class 0 and 93% for class 1 in the training data. Meanwhile, in the testing data, 94% for class 0 and 80% for class 1. Meanwhile, the classification using the Support Vector Machine (SVM) is divided into 3 kernels, namely Radial Basis Function (RBF), Polynomial, and Sigmoid. The best Support Vector Machine (SVM) model is the Support Vector Machine with a Sigmoid kernel with parameter values and C=100γ=0,01 has an F1-Score value of 95% for class 0 and 92% for class 1 on the training data. Meanwhile, in the testing data , 89% for class 0 and 71% for class 1. ANN shows good consistency between F1-Score on training and testing data with small differences. Although SVM with Sigmoid kernel shows a good F1-Score value , ANN excels in data testing.

Keywords: Artificial Neural Network, Financial Distress, F1-Score, Modified Altman, Support Vector Machine.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Kata kunci: Altman Modifikasi, Artificial Neural Network, F1-Score, Financial Distress, Support Vector Machine, Artificial Neural Network, Financial Distress, F1-Score, Modified Altman, Support Vector Machine.
Subjects:	H Social Sciences > HJ Public Finance Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > Q Science (General) > Q325.78 Back propagation Q Science > QA Mathematics > QA353.K47 Kernel functions (analysis) Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Actuaria > 94203-(S1) Undergraduate Thesis
Depositing User:	Dina Silmi Apriliani
Date Deposited:	01 Aug 2024 02:30
Last Modified:	01 Aug 2024 02:30
URI:	http://repository.its.ac.id/id/eprint/110220

Actions (login required)

View Item