Claim Missing Document
Check
Articles

Found 1 Documents
Search

PREDIKSI KONSUMSI ENERGI MENGGUNAKAN REGRESI LINIER DAN MULTILAYER PERCEPTRON (MLP) PADA ALGORITMA KONSENSUS PROOF OF WORK (POW) Wahid Ivan Saputra; Bagus Satrio Waluyo Poetro
Jurnal Rekayasa Sistem Informasi dan Teknologi Vol. 3 No. 1 (2025): Agustus
Publisher : Yayasan Nuraini Ibrahim Mandiri

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.70248/jrsit.v3i1.3000

Abstract

This study discusses energy consumption prediction in blockchain networks based on the Proof of Work (PoW) consensus algorithm, using a Bitcoin case study. The main issue raised is the high energy consumption in the PoW mechanism, which is dynamic and non-linear, making it difficult to accurately predict using simple linear methods. This study proposes a combined approach between the Prophet model to predict energy consumption trends and the Multi-Layer Perceptron (MLP) model to estimate the energy consumption gap (gap_twh) between the estimated and minimum values. The data used comes from Digiconomist, covering the period 2017–2025, with research stages including data cleaning, feature engineering, normalization, modeling, and evaluation using MAE, MSE, and R² metrics. The results show that the combination of Prophet and MLP is able to provide more precise predictions than the linear regression model as a baseline, with interactive visualization through a Streamlit-based dashboard that facilitates interpretation of trends and uncertainty ranges. These findings are expected to serve as a reference for researchers, industry players, and policymakers in monitoring and optimizing energy consumption in crypto mining activities.