Forecasting of individual electricity consumption using Optimized Gradient Boosting Regression with Modified Particle Swarm Optimization

Sepulveda, Luis Fernando Marin; Diniz, Petterson Sousa; Diniz, João Otávio Bandeira; Netto, Stelmo Magalhães Barros; Cipriano, Carolina L. S.; Araújo, Alexandre Cristian; Lemos, Victor Henrique Bezerra de; Pessoa, Alexandre; Quintanilha, Darlan B. P.; Almeida, João Dallyson Sousa de; Silva, Aristófanes Corrêa; Paiva, Anselmo Cardoso de; Bráz, Geraldo; Silva, Márcia I. A.; Monteiro, Eliana Márcia Garros; Silva, Italo Francyles Santos da; Fernandes, Eduardo C.

doi:10.1016/j.engappai.2021.104440

Cited by 21 publications

(2 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…GB models surpass other ML algorithms (such as KNN and SVMs) in accuracy and robustness [74]. A method for predicting energy usage in a large-scale scenario using GB is presented [75].…”

Section: Tree-based and Boosting Modelsmentioning

confidence: 99%

Optimizing Building Short-Term Load Forecasting: A Comparative Analysis of Machine Learning Models

Koukaras,

Mustapha,

Mystakidis

et al. 2024

Energies

View full text Add to dashboard Cite

The building sector, known for its high energy consumption, needs to reduce its energy use due to rising greenhouse gas emissions. To attain this goal, a projection for domestic energy usage is needed. This work optimizes short-term load forecasting (STLF) in the building sector while considering several variables (energy consumption/generation, weather information, etc.) that impact energy use. It performs a comparative analysis of various machine learning (ML) models based on different data resolutions and time steps ahead (15 min, 30 min, and 1 h with 4-step-, 2-step-, and 1-step-ahead, respectively) to identify the most accurate prediction method. Performance assessment showed that models like histogram gradient-boosting regression (HGBR), light gradient-boosting machine regression (LGBMR), extra trees regression (ETR), ridge regression (RR), Bayesian ridge regression (BRR), and categorical boosting regression (CBR) outperformed others, each for a specific resolution. Model performance was reported using R2, root mean square error (RMSE), coefficient of variation of RMSE (CVRMSE), normalized RMSE (NRMSE), mean absolute error (MAE), and execution time. The best overall model performance indicated that the resampled 1 h 1-step-ahead prediction was more accurate than the 15 min 4-step-ahead and the 30 min 2-step-ahead predictions. Findings reveal that data preparation is vital for the accuracy of prediction models and should be model-adjusted.

show abstract

“…GB models surpass other ML algorithms (such as KNN and SVMs) in accuracy and robustness [74]. A method for predicting energy usage in a large-scale scenario using GB is presented [75].…”

Section: Tree-based and Boosting Modelsmentioning

confidence: 99%

Optimizing Building Short-Term Load Forecasting: A Comparative Analysis of Machine Learning Models

Koukaras,

Mustapha,

Mystakidis

et al. 2024

Energies

View full text Add to dashboard Cite

show abstract

“…Gradient boosted regression (GBR) is a synthetic algorithm that develops an enhanced predictor using boosting methods (Wang and Mamo 2020). GBR is inspired by gradient descent, and it approaches the boosting issue as an optimization problem, employing a loss function and attempting to reduce mistakes (Sepulveda et al 2021). In addition to being resilient to outliers, the GBR model can handle a variety of data formats (Wang and Mamo 2020).…”

Section: Selection and Calculation Of The Landscape Indexmentioning

confidence: 99%

Forecasting the effects of habitat fragmentation on the distribution of Sillago larvae and juveniles: a case study in Ka Long Estuary, Vietnam

Thi

Ashley

et al. 2023

Preprint

View full text Add to dashboard Cite

Habitat fragmentation potentially impacts animal diversity due to a decrease in ecological connectivity, especially in estuarine and mangrove environments, where many aquatic organisms spend their early developmental stages on. This study used collections of different species of Sillago larvae and juveniles from 2014 to 2015 in the Ka Long estuary, northern Vietnam to examine the above case based on deep learning models. The combined model, particle swarm optimization - gradient boosting regression, can be used to predict the spatial distribution of Sillago fish in the context of habitat fragmentation, one of the most important parameters to influence in the fish distribution. It reveals that larvae and juveniles of Sillago spp. are more abundant in areas next to mangroves. Based on deep convolutional neural network approach, land use/land cover changed considerably in 2013 and 2023 as mangrove forests and waterbody decreased due to the increase of urbanization. In addition, landscape indices additionally show that the habitat connectivity is decreasing and the number of edges is growing, which indirectly impact the fish larvae and juveniles inhabiting mangroves. Based on genetic algorithm analysis, the present work forecasts Sillago larval and juvenile distribution under the effect of landscape fragmentation in 2033 and 2043, when the Sillago density will have fallen dramatically. Yet, they will choose to reside in mangrove forests, indicating that conserving this environment is essential to maintaining this commercial fish in the study site as well as in adjacent mangrove forests in northern Vietnam.

show abstract

Comparative study of continuous hourly energy consumption forecasting strategies with small data sets to support demand management decisions in buildings

Mariano-Hernández

Hernández-Callejo

Solís

et al. 2022

Energy Science & Engineering

View full text Add to dashboard Cite

Buildings are one of the largest consumers of electrical energy, making it important to develop different strategies to help to reduce electricity consumption. Building energy consumption forecasting strategies are widely used to support demand management decisions, but these strategies require large data sets to achieve an accurate electric consumption forecast, so they are not commonly used for buildings with a short history of record keeping. Based on this, the objective of this study is to determine, through continuous hourly electricity consumption forecasting strategies, the amount of data needed to achieve an accurate forecast. The proposed forecasting strategies were evaluated with Random Forest, eXtreme Gradient Boost, Convolutional Neural Network, and Temporal Convolutional Network algorithms using 4 years of electricity consumption data from two buildings located on the campus of the University of Valladolid.For performance evaluation, two scenarios were proposed for each of the proposed forecasting strategies. The results showed that for forecasting horizons of 1 week, it was possible to obtain a mean absolute percentage error (MAPE) below 7% for Building 1 and a MAPE below 10% for Building

show abstract

Forecasting of individual electricity consumption using Optimized Gradient Boosting Regression with Modified Particle Swarm Optimization

Cited by 21 publications

References 16 publications

Optimizing Building Short-Term Load Forecasting: A Comparative Analysis of Machine Learning Models

Optimizing Building Short-Term Load Forecasting: A Comparative Analysis of Machine Learning Models

Forecasting the effects of habitat fragmentation on the distribution of Sillago larvae and juveniles: a case study in Ka Long Estuary, Vietnam

Comparative study of continuous hourly energy consumption forecasting strategies with small data sets to support demand management decisions in buildings

Contact Info

Product

Resources

About