Rayhan, Ezar Alvah (2025) Model Deteksi Dini Financial Distress pada Perusahaan Sektor Transportasi dan Logistik di Indonesia Menggunakan Metode Explainable Historical Random Forest dan Extreme Gradient Boosting. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 2043211090-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
2043211090-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy |
Abstract
Sektor transportasi dan logistik, yang merupakan kunci penting dalam pembangunan ekonomi Indonesia, menghadapi tantangan yang cukup serius beberapa tahun terakhir. Salah satunya adalah Pandemi COVID-19 pada tahun 2020 yang berdampak pada performa keuangan sektor transportasi dan logistik hingga setelah beberapa tahun berlalu, sektor transportasi dan logistik dinilai masih mengalami permasalahan pada performa keuangannya. Berdasarkan permasalahan tersebut, perusahaan sektor transportasi dan logistik dituntut untuk lebih memerhatikan kondisi keuangannya agar tidak kehilangan kepercayaan investor yang berperan penting pada pendanaan operasional perusahaan. Untuk itu, dilakukan penelitian ini yang akan berfokus dalam membangun Model Deteksi Dini Financial Distress pada Perusahaan Sektor Transportasi dan Logistik Menggunakan Metode Historical Random Forest dan Extreme Gradient Boosting berdasarkan analisis rasio keuangan meliputi rasio likuiditas, aktivitas, profitabilitas, solvabilitas, dan rasio pasar yang diduga memengaruhi terjadinya financial distress. Penelitian ini menunjukkan bahwa metode Historical Random Forest dengan SMOTE mampu menghasilkan akurasi sebesar 97,73%, sensitivitas 100%, spesifisitas 97,06%, dan AUC sebesar 99,71%. Sementara itu, metode Extreme Gradient Boosting dengan SMOTE memberikan hasil terbaik dengan akurasi, sensitivitas, spesifisitas, dan AUC mencapai 100%. Variabel Market to Book Value (X11) dan Debt to Equity Ratio (X9) diidentifikasi sebagai variabel yang paling berpengaruh terhadap financial distress. Untuk mempermudah pengguna dalam mengakses hasil analisis dalam penelitian ini, dibuatlah sebuah Dashboard Interaktif yang mampu melakukan analisis rasio keuangan, prediksi financial distress dan analisis tingkat kepentingan variabel.
========================================================================================================================================
The transportation and logistics sector, which is a key driver of Indonesia's economic development, has faced significant challenges in recent years. One of the major challenges was the COVID-19 pandemic in 2020, which severely impacted the financial performance of the sector. Even years after the pandemic, the transportation and logistics sector is still considered to have unresolved financial performance issues. In light of these challenges, companies in this sector are required to pay closer attention to their financial condition to maintain investor confidence, which is crucial for operational funding. To address this issue, this study focuses on developing an Early Detection Model for Financial Distress in the Transportation and Logistics Sector using the Historical Random Forest and Extreme Gradient Boosting methods. The analysis is based on financial ratios, including liquidity, activity, profitability, solvency, and market ratios, which are hypothesized to influence financial distress. The study reveals that the Historical Random Forest method with SMOTE achieves an accuracy of 97.73%, sensitivity of 100%, specificity of 97.06%, and an AUC of 99.71%. Meanwhile, the Extreme Gradient Boosting method with SMOTE delivers the best results, achieving accuracy, sensitivity, specificity, and an AUC of 100%. The variables Market to Book Value (X11) and Debt to Equity Ratio (X9) are identified as the most influential factors affecting financial distress. To make the findings of this study more accessible to users, an Interactive Dashboard has been developed. This dashboard facilitates financial ratio analysis, financial distress prediction, and variable importance analysis.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Financial Distress, Dashboard, Extreme Gradient Boosting, Historical Random Forest, SMOTE, Transportasi & Logistik, Variable Importance,Transportation & Logistic |
| Subjects: | Q Science Q Science > Q Science (General) Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. |
| Divisions: | Faculty of Vocational > 49501-Business Statistics |
| Depositing User: | Ezar Alvah Rayhan |
| Date Deposited: | 31 Jul 2025 08:08 |
| Last Modified: | 31 Jul 2025 08:08 |
| URI: | http://repository.its.ac.id/id/eprint/125232 |
Actions (login required)
 |
View Item |