Ng, Jiun Shen (2022) Predicting energy consumption pattern based on top trending videos YouTube 2021 using machine learning techniques. Final Year Project, UTAR.
Abstract
This project is about predicting energy consumption patterns based on trending videos on YouTube 2021 by using machine learning techniques. It is important to be aware of how much energy has been consumed while an individual is streaming or watching a video on YouTube on a mobile device. This is because the users will learn how much energy they have wasted as a result of these actions, as well as what the company should do to limit the amount of energy lost. This project will cover the methodology, concept, and design of utilising machine learning language to anticipate energy consumption patterns. The process of putting a machine learning pipeline into action will be explored and analysed. The flow of this project will be presented as follows: exploring the data, pre-processing, model selection and building the model, model validation, fine-tuning the model, testing performance, and finally, simulating it. Jupyter notebook was chosen as the tool for predicting the model, and Python was chosen as the coding language. To help collect the data, AccuBattery was chosen to calculate the energy consumption per video. In this project, several models will be presented and analysed, with the normal equation in Linear Regression will be the algorithm used to simulate it. Furthermore, the data will be clustered in this project utilising threshold-based approaches. Streamlit.io application was chosen as the website to represent the coding. At the end of this project, a formula equation and graph will be thoroughly discussed and analysed for projecting energy consumption patterns based on the most trending videos on YouTube in 2021.
Actions (login required)
![[img]](http://eprints.utar.edu.my/style/images/fileicons/application_pdf.png)
