Anomaly Detection in Gas Pipes with an Ensemble Learning Approach: Combination of Random Forest and GBoost - Dalam bentuk buku karya ilmiah
NOVALDI RAMADHAN WALUYO

Informasi Umum

Kode

24.04.5386

Klasifikasi

000 - General Works

Jenis

Karya Ilmiah - Skripsi (S1) - Reference

Subjek

Data Science

Dilihat

109 kali

Informasi Lainnya

Abstraksi

<p><b>Gas pipeline networks are essential for the safe</b></p>

<p><b>and efficient distribution of gas to various locations, but they are</b></p>

<p><b>also vulnerable to numerous technical issues, with gas leaks</b></p>

<p><b>being one of the most dangerous. Gas leaks in pipelines can lead</b></p>

<p><b>to catastrophic outcomes, including fires, explosions, and</b></p>

<p><b>significant environmental harm. Early detection of these leaks is</b></p>

<p><b>therefore crucial to prevent such severe consequences. This</b></p>

<p><b>research focuses on developing a robust anomaly detection</b></p>

<p><b>method for gas pipeline networks using an ensemble-based</b></p>

<p><b>machine learning approach, specifically through random forest</b></p>

<p><b>and gradient boosting algorithms. The study highlights the</b></p>

<p><b>critical importance of early detection of gas leaks in pipeline</b></p>

<p><b>infrastructure to prevent catastrophic consequences, including</b></p>

<p><b>fires, explosions, and environmental damage. Leveraging</b></p>

<p><b>extensive operational pipeline datasets from oil and gas</b></p>

<p><b>companies, the research begins with a comprehensive data</b></p>

<p><b>preprocessing phase designed to ensure the highest level of data</b></p>

<p><b>quality and integrity. Both random forest and gradient boost</b></p>

<p><b>models are rigorously implemented and trained on this dataset,</b></p>

<p><b>with a focus on clustering data into decision trees or groups to</b></p>

<p><b>effectively identify anomalies. The primary objective is to</b></p>

<p><b>compare the accuracy of the random forest and gradient boost</b></p>

<p><b>models while also exploring the potential for enhanced</b></p>

<p><b>performance by combining these two powerful methods. The</b></p>

<p><b>effectiveness of the anomaly detection system is meticulously</b></p>

<p><b>evaluated using F1-score and accuracy metrics, which provide a</b></p>

<p><b>clear measure of model performance. This research aims to</b></p>

<p><b>significantly improve the safety and reliability of gas</b></p>

<p><b>distribution systems by delivering a cutting-edge machine</b></p>

<p><b>learning approach for anomaly detection in gas pipelines. The</b></p>

<p><b>study's results, demonstrating an accuracy of 0.90 and an F1-</b></p>

<p><b>score of 0.90, indicate strong and reliable performance.</b></p>

Koleksi & Sirkulasi

Tersedia 1 dari total 1 Koleksi

Pengarang

Penerbit

Sirkulasi

Download / Flippingbook