Large neural networks for electricity load forecasting: are they overfitted?

Article Abstract:

Neural networks are used for forecasting daily electricity loads and demand. The efficacy of large neural network specifications and consequent over-fitting issues for forecasting purposes are examined by comparing different forecasting methods for hourly electricity demand. Neural network method is also compared with conventional regression-based method.

United Kingdom

---

Forecasting electricity prices for a day-ahead pool-based electric energy market

Article Abstract:

Methods for forecasting prices of a day-ahead electric energy market are compared. The methods examined include time series analysis, neural networks and wavelets.

Commodity & service prices, Spain, Prices and rates, Time-series analysis, Time series analysis, Company pricing policy

---

Forecasting electricity demand using generalized long memory

Article Abstract:

A stochastic model, using generalized long memory, is presented for studying the seasonal patterns of hourly electricity load demand in Brazil.

Learning models (Stochastic processes), Stochastic models

Similar abstracts:

• Abstracts: Monetary and fiscal remedies for deflation. Dividend taxes and firm valuation: new evidence. Capital-gains realizations of the rich and sophisticated
• Abstracts: Author's retrospective on 'forecasting seasonals and trends by exponentially weighted moving averages'. Charles Holt's report on exponentially weighted moving averages: an introduction and appreciation
• Abstracts: Decomposition by causal forces: a procedure for forecasting complex time series. Forecasting support systems for the incorporation of event information: an empirical investigation

This website is not affiliated with document authors or copyright owners. This page is provided for informational purposes only. Unintentional errors are possible.