Sparse Recurrent Mixture Density Networks for Forecasting High Variability Time Series with Confidence Estimates.
Narendhar GugulothuEaswar SubramanianSanjay P. BhatPublished in: ICANN (2) (2019)
Keyphrases
- confidence estimates
- weather forecasting
- financial time series
- forecasting accuracy
- arma model
- recurrent networks
- short term
- exponential smoothing
- box jenkins
- stock market
- high dimensional
- dynamic time warping
- sparse data
- social networks
- mixture model
- recurrent neural networks
- non stationary
- heavy tailed distributions
- chaotic time series
- complex networks
- arima model
- multi variate
- software defect
- spiking neural networks
- moving average
- multivariate time series
- compressive sensing
- hybrid model
- support vector regression
- gaussian distribution
- feed forward
- network structure
- long term