A Blockwise Embedding for Multi-Day-Ahead Prediction of Energy Time Series by Randomized Deep Neural Networks.
Francesco Di LuzioAntonello RosatoFederico SuccettiMassimo PanellaPublished in: IJCNN (2021)
Keyphrases
- neural network
- short term prediction
- short term
- chaotic time series
- mackey glass
- long term
- financial time series
- artificial neural networks to predict
- prediction accuracy
- moving average
- prediction model
- exponential smoothing
- pattern recognition
- wind power
- back propagation
- energy consumption
- recurrent neural networks
- non stationary
- box jenkins
- stock market
- energy minimization
- prediction algorithm
- wind speed
- stochastic programming
- power generation
- extreme values
- neural network ensemble
- vector space
- artificial neural networks
- multivariate time series
- radial basis function
- predictive model
- linear program
- dynamic time warping
- prediction error
- neural network model