Loss-Customised Probabilistic Energy Time Series Forecasts Using Automated Hyperparameter Optimisation.
Kaleb PhippsStefan MeisenbacherBenedikt HeidrichMarian TurowskiRalf MikutVeit HagenmeyerPublished in: e-Energy (2023)
Keyphrases
- probabilistic model
- data driven
- energy consumption
- energy minimization
- context sensitive
- semi automated
- gaussian processes
- dynamic time warping
- non stationary
- short term
- genetic algorithm
- wind farm
- historical data
- uncertain data
- long term
- energy function
- fully automated
- generative model
- model selection
- bayesian networks
- neural network
- discrete valued
- forecast models