Prediksi Financial Distress Pada Perusahaan Sektor Transportasi di Indonesia Menggunakan Support Vector Machine

Yulianto, Ahmat (2023) Prediksi Financial Distress Pada Perusahaan Sektor Transportasi di Indonesia Menggunakan Support Vector Machine. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06211940000083-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (2MB) | Request a copy

Abstract

Industri transportasi di Indonesia, baik sebagai infrastruktur maupun sebagai layanan, berperan penting dalam aktivitas ekonomi dan penentuan daya saing ekonomi suatu negara. Namun, menurut data dari BPS (2021), sektor transportasi pergudangan merupakan sektor yang paling terkena dampak pandemi Covid-19. Sehingga terdapat ancaman perusahaan transportasi akan mengalami kebangkrutan. Sebelum mengalami kebangkrutan, yang dialami oleh perusahaan adalah financial distress. Financial distress adalah tahap dimana suatu perusahaan mengalami penurunan kondisi keuangan ataupun likuidasi. Penelitian ini bertujuan untuk memprediksi financial distress pada perusahaan transportasi di Indonesia yang terdaftar di Bursa Efek Indonesia menggunakan Support Vector Machine (SVM). Data yang digunakan adalah data 45 perusahaan sektor transportasi di Indonesia tahun 2021. Variabel independen dan dependen yang digunakan pada penlitian ini secara berturut-turut adalah 20 rasio keuangan dan 1 variabel klasifikasi biner (financial distress = 1 atau non financial distress = 0) yang diklasifikasikan melalui model Altman. Pembagian data training dan data testing menggunakan stratified k-fold dengan jumlah fold adalah 10 fold. Model SVM terbaik dicari menggunakan fitur GridSearchCV pada software python pada setiap fold. Parameter model SVM terbaik adalah parameter yang paling sering muncul pada fold 1 sampai 10, yaitu dengan nilai C = 5 dan y = 1 yang menjadi parameter terbaik pada fold 3,4,6 dan 8. Kemudian model dilakukan pelatihan, dari hasil latih model SVM terbaik, diperoleh nilai loss function pada fold 1 sampai dengan fold 10 secara berturut-turut adalah sebagai berikut 0,126, 0,105, 0,193, 0,208, 0,363, 0,187, 0,001, 0,224, 0,040, 0,107, yang artinya model memiliki tingkat kesalahan yang cukup rendah dalam mengklasifikasikan kelas data. Metrik evaluasi rata-rata dari cross-validation mengungkap performa model machine learning yang telah dilatih dengan 10-fold cross-validation pada data training. Rata-rata accuracy, precission, recall, F1-Score dan AUC secara berturut-turut adalah 0,664, 0,549, 0,7, 0,563, 0,683.
=================================================================================================================================
The transportation industry in Indonesia, both as infrastructure and as a service, plays an important role in economic activity and determines a country's economic competitiveness. However, according to data from BPS (2021), the warehousing transportation sector is the sector most affected by the Covid-19 pandemic. So there is a threat that the transportation company will go bankrupt. Before experiencing bankruptcy, what the company experienced was financial distress. Financial distress is the stage where a company experiences a decline in financial condition or liquidation. This study aims to predict financial distress in transportation companies in Indonesia that are listed on the Indonesia Stock Exchange using a Support Vector Machine (SVM). The data used is data from 45 companies in the transportation sector in Indonesia in 2021. The independent and dependent variables used in this research are respectively 20 financial ratios and 1 binary classification variable (financial distress = 1 or non-financial distress = 0) which are classified as via the Altman model. Distribution of training data and data testing using stratified k-fold with a total of 10 folds. The best SVM model is searched using the GridSearchCV feature in the python software for each fold. The best SVM model parameters are the parameters that appear most often in folds 1 to 10, namely with the values C=5 and y=1 which are the best parameters in folds 3,4,6 and 8. Then the model is trained, from the results of the SVM model training the best, the loss function values obtained at fold 1 to fold 10 are as follows 0.126, 0.105, 0.193, 0.208, 0.363, 0.187, 0.001, 0.224, 0.040, 0.107, which means that the model has a fairly low error rate in classify data classes. The average evaluation metric from cross-validation reveals the performance of a machine learning model that has been trained with 10-fold cross-validation on training data. The average accuracy, precision, recall, F1-Score and AUC were 0.664, 0.549, 0.7, 0.563, 0.683 respectively.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Altman Z-Score, Financial Distress, Nilai Akurasi, Support Vector Machine, Perusahaan Transportasi, Accuracy Value, Transportation Company
Subjects:	H Social Sciences > HA Statistics H Social Sciences > HA Statistics > HA29 Theory and method of social science statistics
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Ahmat Yulianto
Date Deposited:	18 Oct 2023 07:52
Last Modified:	18 Oct 2023 07:52
URI:	http://repository.its.ac.id/id/eprint/104561

Actions (login required)

View Item