Time Series Decomposition for Improving the Forecasting Performance of Convolutional Neural Networks.
Iván Méndez-JiménezMiguel Cárdenas-MontesPublished in: CAEPIA (2018)
Keyphrases
- convolutional neural networks
- weather forecasting
- financial time series
- forecasting accuracy
- short term
- arma model
- dynamic time warping
- convolutional network
- quasi periodic
- box jenkins
- hybrid model
- exponential smoothing
- stock market
- non stationary
- multivariate time series
- arima model
- chaotic time series
- image decomposition
- moving average
- decomposition methods
- decomposition method
- long term
- multiscale
- load forecasting
- data mining
- exchange rate
- phase space reconstruction
- stock price
- support vector regression
- artificial neural networks