Forecasting in the energy sector has developed rapidly in the past few years. Two promising steps to improve forecasting accuracy include applications of new methods and extended use of new data. Models based on methods such as advanced econometric techniques, deep learning, text mining and complex data processing methods have shown some superiority over traditional models. While data such as internet search frequency, social media posts, sensor data and grid data, provide new insights during forecasting process. Research also shows evidence that a combination model outperforms a single model, which leads to model combination being a fast-growing research area.
The objective of this Research Topic is to present cutting edge progresses in energy sector forecasting and inspire new methods for future research. The topic covers forecasting problems in most of the energy sector, including crude oil, natural gas, electricity, and renewable energy. Various aspects of specific energy can be addressed, such as prices, demand, production, inventory, and trade. Connections between different kinds of energy is important too. Special attention should be paid to advanced models and utilization of new data in energy forecasting, such as deep learning and big data. Ensemble forecasting and other new progresses in energy forecasting would be great additions to this Research Topic.
Topics of interest include, but are not limited to:
· Energy price forecasting,
· Energy demand and supply forecasting,
· Energy import and export forecasting,
· Load forecasting,
· Renewable energy forecasting,
· Advanced econometric models,
· Mixed frequency model,
· High frequency data,
· Interval data,
· Sensor data,
· Climate data,
· Grid data,
· Big data,
· Panel data,
· Deep learning,
· Artificial intelligence,
· Text mining,
· Data mining,
· Complex data preprocessing methods,
· Empirical mode decomposition,
· Singular spectrum analysis,
· Direct filter analysis,
· Model combination,
· Model average,
· Ensemble model,
· Forecasting evaluation.
Forecasting in the energy sector has developed rapidly in the past few years. Two promising steps to improve forecasting accuracy include applications of new methods and extended use of new data. Models based on methods such as advanced econometric techniques, deep learning, text mining and complex data processing methods have shown some superiority over traditional models. While data such as internet search frequency, social media posts, sensor data and grid data, provide new insights during forecasting process. Research also shows evidence that a combination model outperforms a single model, which leads to model combination being a fast-growing research area.
The objective of this Research Topic is to present cutting edge progresses in energy sector forecasting and inspire new methods for future research. The topic covers forecasting problems in most of the energy sector, including crude oil, natural gas, electricity, and renewable energy. Various aspects of specific energy can be addressed, such as prices, demand, production, inventory, and trade. Connections between different kinds of energy is important too. Special attention should be paid to advanced models and utilization of new data in energy forecasting, such as deep learning and big data. Ensemble forecasting and other new progresses in energy forecasting would be great additions to this Research Topic.
Topics of interest include, but are not limited to:
· Energy price forecasting,
· Energy demand and supply forecasting,
· Energy import and export forecasting,
· Load forecasting,
· Renewable energy forecasting,
· Advanced econometric models,
· Mixed frequency model,
· High frequency data,
· Interval data,
· Sensor data,
· Climate data,
· Grid data,
· Big data,
· Panel data,
· Deep learning,
· Artificial intelligence,
· Text mining,
· Data mining,
· Complex data preprocessing methods,
· Empirical mode decomposition,
· Singular spectrum analysis,
· Direct filter analysis,
· Model combination,
· Model average,
· Ensemble model,
· Forecasting evaluation.