Prediksi Pendapatan Perusahaan Berdasarkan ESG Risk Rating Score untuk Kebutuhan Sustainable Finance Menggunakan Algoritma XGBoost

Ramadhan, Prima Secondary (2023) Prediksi Pendapatan Perusahaan Berdasarkan ESG Risk Rating Score untuk Kebutuhan Sustainable Finance Menggunakan Algoritma XGBoost. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 05311940000001-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2025. Download (1MB) | Request a copy

Abstract

Dalam sesi Leader's Insight, Gubernur Perry Warjiyo menyatakan Sustainable Finance Working Group G20 telah saling bekerja sama dengan fokus pengembangan Sustainable Finance Instrument (SFI), termasuk mengatasi berbagai tantangan penerapannya, untuk mendukung ekonomi hijau dan berkelanjutan. Sustainable finance mengacu pada proses mempertimbangkan environmental, social dan governance (ESG) ketika membuat keputusan investasi di sektor keuangan, yang mengarah pada investasi jangka panjang yang lebih banyak dalam kegiatan dan proyek ekonomi berkelanjutan. Bisnis saat ini saling berhubungan secara global. Pemangku kepentingan menyadari bahwa tanggung jawab ESG suatu perusahaan merupakan bagian integral dari kinerja dan keberlanjutan jangka panjangnya. Oleh karena itu, pada Tugas Akhir ini dilakukan penelitian terhadap hubungan antara ESG Risk Rating Score dengan kekayaan perusahaan seperti; employees, profits, assets, dan market value. Tahap pertama dalam penelitian ini adalah persiapan data, tahap kedua penentuan model machine learning terbaik menggunakan library lazypredict, tahap ketiga membangun model berdasarkan saran pada tahap kedua, tahap keempat hyperparameter tuning model menggunakan library GridSearchCV, dan tahap kelima adalah model evaluasi menggunakan R2, MSE, RMSE, dan MAE. Pada tahap persiapan data, diterapkan data splitting dan normalization pada 920 data yang didapatkan dari Fortune dan Sustainalitycs. Tahap berikutnya, dilakukan run library lazypredict yang menghasilkan output model terbaik yaitu ExtraTrees dengan R2 sebesar 0,8732. Diikuti algoritma yang diusulkan oleh penulis, XGBoost, yang hanya mendapatkan R2 sebesar 0,7489. Tahap ketiga dan keempat adalah hyperparmeter tuning dan model evaluasi, dimana dilakukan run library SigOpt yang dapat meningkatkan akurasi R2 dari model ExtraTrees dan XGBoost secara berurutan menjadi 0,8838 dan 0,8537.
=================================================================================================================================
In the Leader's Insight session, Governor Perry Warjiyo stated that the G20 Sustainable Finance Working Group had collaborated with a focus on developing the Sustainable Finance Instrument (SFI), including overcoming various challenges in its implementation, to support a green and sustainable economy. Sustainable finance refers to the process of considering environmental, social, and governance (ESG) when making investment decisions in the financial sector, leading to more long-term investment in sustainable economic activities and projects. Today's business is interconnected globally. Stakeholders recognize that a company's ESG responsibility is an integral part of its long-term performance and sustainability. As a result of these conditions, the researcher wants to conduct a study on the relationship between ESG Risk Rating Score and company assets such as; the number of employees, total profits, assets, and market value. The first stage of this research is data preparation, the second stage is determining the best machine learning model using the lazypredict library, the third stage is building a model based on suggestions in the second stage, the fourth stage is hyperparameter tuning model using the GridSearchCV library, and the fifth stage is an evaluation model using R2, MSE, RMSE, and MAE. In the data preparation stage, data splitting and normalization were applied to 920 rows of data obtained from Fortune and Sustainalitycs. The next step is to run the lazypredict library, which produces the best model output, namely ExtraTrees with an R2 of 0.8732. Followed by the algorithm proposed by the author, XGBoost, which only gets an R2 of 0.7489. The third and fourth stages are hyperparameter tuning and model evaluation, where the SigOpt run library is carried out, which can increase the R2 accuracy of the ExtraTrees and XGBoost models, respectively, to 0.8838 and 0.8537.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Presidensi G20, Sustainable Finance, ESG, Multiple Linear Regression, Lazypredict Library, XGBoost, G20 Presidency, Sustainable Finance, ESG, Multiple Linear Regression, Lazypredict Library, XGBoost.
Subjects:	T Technology > T Technology (General) > T174 Technological forecasting T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T57.74 Linear programming
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis
Depositing User:	Prima Secondary Ramadhan
Date Deposited:	11 Feb 2023 10:30
Last Modified:	11 Feb 2023 10:30
URI:	http://repository.its.ac.id/id/eprint/96744

Actions (login required)

View Item