Skripsi

KLASIFIKASI AKTIVITAS BOT MAXIMAL EXTRACTABLE VALUE (MEV) PADA TRANSAKSI BLOCKCHAIN MENGGUNAKAN MODEL EXTREME GRADIENT BOOSTING

---

Maximal Extractable Value (MEV) bot activity on blockchain networks poses a significant challenge, as MEV bots exploit transaction-processing mechanisms to gain profit in ways that may hinder fairness, increase gas fees, and disrupt network stability. This study employs the Extreme Gradient Boosting (XGBoost) model to classify MEV bot activity in Ethereum blockchain transactions using numerical and categorical features extracted from historical data. The research workflow includes initial data analysis, correlation analysis, elimination of features with high multicollinearity, and standardization using the Z-Score method, followed by model training across four data-splitting scenarios and hyperparameter tuning on n_estimators, learning_rate, and max_depth. Experimental results show that XGBoost achieves highly accurate classification across all scenarios, with accuracy exceeding 0.99, and the best performance obtained under the 80:10:10 split with 0.9977 accuracy, 0.9975 precision, 0.9982 recall, and 0.9979 F1-score. These findings indicate that XGBoost is effective, stable, and efficient for classifying MEV bot activity. Future work may explore alternative models such as LightGBM or CatBoost and develop automated real-time monitoring systems for deployment within blockchain networks. Keywords: Blockchain, Ethereum, Networking, MEV, Machine Learning, XGBoost

---

Availability

Detail Information

Series Title

-

Call Number

T1903792025

Publisher

Indralaya : Prodi Sistem Komputer, Fakultas Ilmu Komputer Universitas Sriwijaya., 2025

Collation

xiv, 86 hlm.; lamp.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

006.312 07

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Sistem Komputer
Extreme Gradient Boosting

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment