SSDNet: State Space Decomposition Neural Network for Time Series Forecasting.
Yang LinIrena KoprinskaMashud RanaPublished in: ICDM (2021)
Keyphrases
- state space
- neural network
- artificial neural networks
- state variables
- reinforcement learning
- back propagation
- dynamical systems
- optimal policy
- planning problems
- markov decision processes
- heuristic search
- dynamic programming
- search space
- decomposition method
- particle filter
- multi layer
- hidden layer
- bp neural network
- genetic algorithm
- self organizing maps
- fuzzy logic
- neural network model
- multilayer perceptron
- recurrent neural networks
- prediction model
- basis functions
- fuzzy neural network
- multi layer perceptron
- neural network is trained
- feed forward neural networks
- activation function
- continuous time markov process
- image reconstruction from projections
- continuous state spaces
- auto associative
- markov decision problems
- decomposition methods
- multiscale
- reinforcement learning algorithms
- search algorithm
- radial basis function