Prediction Interval Adjustment for Load-Forecasting using Machine Learning

Zuniga-Garcia, Miguel A.; Santamaría-Bonfil, Guillermo; Figueroa, Gustavo; Batres, Rafael

doi:10.3390/app9245269

Cited by 9 publications

(7 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results showed that the ANN model outperformed traditional statistical models in predicting rice yield. However, the researcher concluded that further studies is required to improve the accuracy and interpretability of the model [75][76][77]. While ANNs have the potential to improve agriculture, there are limitations and drawbacks that need to be addressed for better results.…”

Section: Observationmentioning

confidence: 99%

Artificial Neural Networks and their Use in Agricultural Genetic Engineering and Protein Estimation

Kotiyal,

Dudi,

Lakshmi

et al. 2023

IJECC

View full text Add to dashboard Cite

A mathematical model called an artificial neural network uses certain algorithms to anticipate and predict various events. This model has multiple layers, including input, hidden, and output layers. By altering its algorithms, various outputs can be produced based on the input utilized. Biological neuron mechanization has an impact on artificial neural networks. As biological neurons have a propensity to learn and train sets of data for producing biased outputs by spotting and removing variances in them, ANN also works on these principles. Although this model has many uses, it has historically been employed in biological experiments using the supervised learning method, one of which is to predict protein secondary structure. This allows one to identify the positions of different amino acids that are ordered complicated protein structures, which are very appealing in genetic engineering. By this model scientists can map out and isolate a desirable gene in genetic material without going through intense laborious experiments. This report summarizes all the objectives linked to artificial neural networks as well as their applications in bioengineering by examining many related studies.

show abstract

Section: Observationmentioning

confidence: 99%

Artificial Neural Networks and their Use in Agricultural Genetic Engineering and Protein Estimation

Kotiyal,

Dudi,

Lakshmi

et al. 2023

IJECC

View full text Add to dashboard Cite

show abstract

“…This section describes the error metrics used to evaluate the performance of the models. In this work, two error metrics were used to score and compare forecast values to reference values [29,30].…”

Section: Error Metricsmentioning

confidence: 99%

A Data-Driven Approach to Forecasting Heating and Cooling Energy Demand in an Office Building as an Alternative to Multi-Zone Dynamic Simulation

et al. 2021

View full text Add to dashboard Cite

Nowadays, as more data is now available from an increasing number of installed sensors, load forecasting applied to buildings is being increasingly explored. The amount and quality of resulting information can provide inputs for smarter decisions when managing and operating office buildings. In this article, the authors use two data-driven methods (artificial neural networks and support vector machines) to predict the heating and cooling energy demand in an office building located in Lisbon, Portugal. In the present case-study, these methods prove to be an accurate and appealing alternative to the use of accurate but time-consuming multi-zone dynamic simulation tools, which strongly depend on several parameters to be inserted and user expertise to calibrate the model. Artificial neural networks and support vector machines were developed and parametrized using historical data and different sets of exogenous variables to encounter the best performance combinations for both the heating and cooling periods of a year. In the case of support vector regression, a variation introduced simulated annealing to guide the search for different combinations of hyperparameters. After a feature selection stage for each individual method, the results for the different methods were compared, based on error metrics and distributions. The outputs of the study include the most suitable methodology for each season, and also the features (historical load records, but also exogenous features such as outdoor temperature, relative humidity or occupancy profile) that led to the most accurate models. Results clearly show there is a potential for faster, yet accurate machine-learning based forecasting methods to replace well-established, very accurate but time-consuming multi-zone dynamic simulation tools to forecast building energy consumption.

show abstract

“…Another metric that is introduced to measure the error of the constructed prediction intervals is the probabilistic RMSE (p-RMSE) [34]. This metric aims to quantify the error in cases the target values, i.e., the measured actual load, exceed the bounds of the prediction interval that is selected by the planner.…”

Section: Assessment Metrics Of Probabilistic Forecastsmentioning

confidence: 99%

Probabilistic Spatial Load Forecasting Based on Hierarchical Trending Method

Evangelopoulos¹,

Karafotis²,

Georgilakis³

2020

Energies

View full text Add to dashboard Cite

The efficient spatial load forecasting (SLF) is of high interest for the planning of power distribution networks, mainly in areas with high rates of urbanization. The ever-present spatial error of SLF arises the need for probabilistic assessment of the long-term point forecasts. This paper introduces a probabilistic SLF framework with prediction intervals, which is based on a hierarchical trending method. More specifically, the proposed hierarchical trending method predicts the magnitude of future electric loads, while the planners’ knowledge is used to improve the allocation of future electric loads, as well as to define the year of introduction of new loads. Subsequently, the spatial error is calculated by means of root-mean-squared error along the service territory, based on which the construction of the prediction intervals of the probabilistic forecasting part takes place. The proposed probabilistic SLF is introduced to serve as a decision-making tool for regional planners and distribution network operators. The proposed method is tested on a real-world distribution network located in the region of Attica, Athens, Greece. The findings prove that the proposed method shows high spatial accuracy and reduces the spatial error compared to a business-as-usual approach.

show abstract

Prediction Interval Adjustment for Load-Forecasting using Machine Learning

Cited by 9 publications

References 28 publications

Artificial Neural Networks and their Use in Agricultural Genetic Engineering and Protein Estimation

Artificial Neural Networks and their Use in Agricultural Genetic Engineering and Protein Estimation

A Data-Driven Approach to Forecasting Heating and Cooling Energy Demand in an Office Building as an Alternative to Multi-Zone Dynamic Simulation

Probabilistic Spatial Load Forecasting Based on Hierarchical Trending Method

Contact Info

Product

Resources

About