Nature inspired based optimize combination of time series and machine learning model for predict energy consumption of a residence

Energy consumption has gradually increased greenhouse gas concentrations in the atmosphere and is considered the main source of global warming. Therefore, many studies have been developed prediction models with higher accuracy and lower feature cost by the day. This study proposed a time series fore...

Full description

Saved in:
Bibliographic Details
Main Authors: Le, Thuy-Linh, Truong, Dinh-Nhat, Nguyen, Duc Sy
Format: Conference Proceeding
Language:English
Subjects:
Algorithms
Atmospheric models
Energy consumption
Greenhouse effect
Greenhouse gases
Machine learning
Optimization
Prediction models
Support vector machines
Time series
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Energy consumption has gradually increased greenhouse gas concentrations in the atmosphere and is considered the main source of global warming. Therefore, many studies have been developed prediction models with higher accuracy and lower feature cost by the day. This study proposed a time series forecasting model by integrates a new nature-inspired optimization algorithm (Jellyfish Algorithm) with a machine learning model (least squares support vector regression). Notably, the data set will be pre-processing to build the most efficient model and the parameters of machine learning are optimized by Jelly Fish Algorithm. The proposed model was evaluated by a real- time dataset and was compared with other regression, times series models, and previous studies. Results reveal that this model is an effective model to predict energy consumption.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0124782