Interval prediction for time series based on LSTM and mixed Gaussian distribution.
Zongxia XieRenhui LiHui HuPublished in: SMC (2020)
Keyphrases
- gaussian distribution
- multi variate
- maximum likelihood
- financial time series
- noise model
- maximum a posteriori estimation
- prediction accuracy
- gaussian mixture model
- exponential smoothing
- chaotic time series
- prediction error
- box jenkins
- prediction model
- recurrent neural networks
- gaussian model
- non stationary
- moving average
- prediction algorithm
- dynamic time warping
- stock market
- extreme values
- neural network
- intensity distribution
- pairwise
- object recognition
- training data
- clustering algorithm
- learning algorithm