Blending gradient boosted trees and neural networks for point and probabilistic forecasting of hierarchical time series.
Ioannis NasiosKonstantinos VogklisPublished in: CoRR (2023)
Keyphrases
- neural network
- forecasting accuracy
- chaotic time series
- weather forecasting
- mackey glass
- short term prediction
- financial time series
- neural network model
- short term
- non stationary
- exponential smoothing
- decision trees
- gradient vectors
- arma model
- fuzzy logic
- bayesian networks
- pattern recognition
- artificial neural networks
- bp neural network
- probabilistic model
- demand forecasting
- multivariate time series
- financial forecasting
- forecasting model
- uncertain data
- backpropagation neural networks
- multilayer perceptron
- back propagation
- generative model
- grey model
- dynamic time warping
- binary tree
- exchange rate
- hybrid model
- multi layer
- stock market
- long term
- genetic algorithm
- moving average
- gradient information
- garch model
- turning points
- recurrent neural networks
- fault diagnosis
- learning algorithm