Prediction of Ultimate Fighting Championship year [1994-2022] using xgboost and random forest algorithms = Prediksi hasil pertandingan Ultimate Fighting Championship tahun [1994-2022] menggunakan algoritma xgboost dan random forest

Kristanto, Jason (2025) Prediction of Ultimate Fighting Championship year [1994-2022] using xgboost and random forest algorithms = Prediksi hasil pertandingan Ultimate Fighting Championship tahun [1994-2022] menggunakan algoritma xgboost dan random forest. Bachelor thesis, Universitas Pelita Harapan.

[thumbnail of Title]
Preview
Text (Title)
Title.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (146kB) | Preview
[thumbnail of Abstract]
Preview
Text (Abstract)
Abstract.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (348kB) | Preview
[thumbnail of ToC]
Preview
Text (ToC)
ToC.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (599kB) | Preview
[thumbnail of Chapter1]
Preview
Text (Chapter1)
Chapter1.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB) | Preview
[thumbnail of Chapter2] Text (Chapter2)
Chapter2.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB)
[thumbnail of Chapter3] Text (Chapter3)
Chapter3.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (3MB)
[thumbnail of Chapter4] Text (Chapter4)
Chapter4.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB)
[thumbnail of Chapter5] Text (Chapter5)
Chapter5.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (278kB)
[thumbnail of Bibliography]
Preview
Text (Bibliography)
Bibliography.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (429kB) | Preview
[thumbnail of Appendices] Text (Appendices)
Appendices.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (2MB)

Abstract

The Ultimate Fighting Championship (UFC) is the world’s largest mixed martial arts (MMA) organization, featuring fighters from diverse martial arts disciplines. In today’s data-driven landscape, predicting UFC match outcomes has become increasingly important—not only for coaches and analysts but also for the sports betting industry and fans. The main challenge in making such predictions lies in the numerous statistical variables that influence fight outcomes and the complexity of evolving fight data. With the growing availability of historical fight data, machine learning (ML)-based predictive methods offer a promising solution that can provide higher accuracy than traditional statistical or intuition-based approaches. This study aims to develop a predictive model for UFC fight outcomes by comparing the performance of two popular ML algorithms: Random Forest and Extreme Gradient Boosting (XGBoost). The research methodology consists of data preprocessing, feature engineering, skill and form score calculation, model training, hyperparameter tuning, and evaluation using metrics such as accuracy, precision, recall, F1-score, and ROC-AUC. Experimental results show that the XGBoost model achieved a prediction accuracy of 69.1%, slightly outperforming the Random Forest model, which reached 68.64%. The XGBoost model also attained an F1-score of 0.68 and a recall of 0.69, indicating balanced performance. In contrast, Random Forest obtained an F1-score of 0.67. These findings suggest that XGBoost is more effective at handling the complexity of UFC fight data and provides more stable predictions. Therefore, XGBoost is recommended for developing data-driven UFC fight outcome prediction systems. / Ultimate Fighting Championship (UFC) merupakan organisasi olahraga bela diri campuran (Mixed Martial Arts/MMA) terbesar di dunia, yang menyajikan pertarungan antara petarung dari berbagai disiplin seni bela diri. Dalam dunia yang semakin data-driven, prediksi hasil pertandingan UFC menjadi penting tidak hanya bagi pelatih dan analis olahraga, tetapi juga bagi industri taruhan olahraga dan penggemar. Permasalahan utama dalam prediksi ini adalah banyaknya variabel statistik yang saling mempengaruhi dan kompleksitas data pertarungan yang terus berkembang. Dengan meningkatnya volume data historis dan statistik pertandingan, metode prediktif berbasis pembelajaran mesin (machine learning) menjadi solusi yang potensial untuk menghasilkan prediksi yang lebih akurat dibandingkan metode konvensional. Penelitian ini bertujuan untuk membangun model prediksi hasil pertandingan UFC dengan membandingkan performa dua algoritma machine learning populer, yaitu Random Forest dan Extreme Gradient Boosting (XGBoost). Tahapan penelitian meliputi preprocessing data, feature engineering, pembuatan fitur skor keterampilan dan performa, pelatihan model, tuning hyperparameter, serta evaluasi menggunakan metrik seperti akurasi, precision, recall, F1-score, dan ROC-AUC. Penelitian ini juga melakukan analisis fitur penting yang berkontribusi terhadap prediksi hasil pertandingan. Hasil eksperimen menunjukkan bahwa model XGBoost mampu menghasilkan akurasi prediksi sebesar 69,1%, sedikit lebih tinggi dibandingkan Random Forest yang mencapai 68,64%. Selain itu, XGBoost memiliki skor F1 sebesar 0,68 dengan recall 0,69, yang menunjukkan kinerja model yang seimbang antara sensitivitas dan presisi. Sementara Random Forest menghasilkan skor F1 sebesar 0,67. Temuan ini menunjukkan bahwa XGBoost memiliki keunggulan dalam mengolah kompleksitas data pertarungan UFC dan mampu memberikan prediksi yang lebih stabil. Dengan demikian, model berbasis XGBoost direkomendasikan untuk digunakan dalam sistem prediksi hasil pertandingan UFC berbasis data statistik.
Item Type: Thesis (Bachelor)
Creators:
Creators
NIM
Email
ORCID
Kristanto, Jason
NIM01082180038
jasonkriss13@gmail.com
UNSPECIFIED
Contributors:
Contribution
Contributors
NIDN/NIDK
Email
Thesis advisor
Lazarusli, Irene Astuti
NIDN0317097501
irene.lazarusli@uph.edu
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: University Subject > Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Informatics
Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Informatics
Depositing User: Jason Kristanto
Date Deposited: 28 Jun 2025 05:14
Last Modified: 01 Jul 2025 07:56
URI: http://repository.uph.edu/id/eprint/69056

Actions (login required)

View Item
View Item