Skripsi

PREDIKSI VOLUME TOTAL TRANSAKSI HARIAN ETHERIUM MENGGUNAKAN METODE XGBOOST

---

The development of blockchain technology has increased cryptocurrency transaction activities, especially Ethereum, whose daily transaction volume is highly fluctuating. This condition makes transaction volume prediction important for understanding network activity patterns. This study aims to develop a prediction model for daily Ethereum transaction volume using the XGBoost Regressor algorithm. The dataset was obtained from public Ethereum transaction data on Google BigQuery for the 2020–2024 period. The preprocessing stages included daily aggregation, log1p transformation, and feature engineering consisting of lag features, rolling mean, rolling standard deviation, and calendar features. The model was tested using two scenarios: baseline and hyperparameter tuning with GridSearchCV and TimeSeriesSplit. Performance evaluation used Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R-Squared (R²). The baseline model achieved MAE 0.231, MSE 0.132, RMSE 0.363, and R² 0.952. Meanwhile, the hyperparameter tuning model performed better with MAE 0.191, MSE 0.082, RMSE 0.287, and R² 0.970. These results indicate that hyperparameter tuning improves model performance.

---

Availability

Detail Information

Series Title

-

Call Number

T2012582026

Publisher

Indralaya : Prodi Teknik Informatika, Fakultas Ilmu Komputer Universitas Sriwijaya., 2026

Collation

xv, 99 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

006.307

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Teknik Informatika
Algoritma Prediksi--XGBOOST

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment