Heterogeneity-Informed Meta-Parameter Learning for Spatiotemporal Time Series Forecasting.
Zheng DongRenhe JiangHaotian GaoHangchen LiuJinliang DengQingsong WenXuan SongPublished in: KDD (2024)
Keyphrases
- parameter learning
- bayesian networks
- structure learning
- conditional random fields
- statistical learning
- generative model
- maximum likelihood
- em algorithm
- markov random field
- parameter estimation
- expectation maximization
- approximate inference
- moving objects
- probabilistic model
- least squares
- machine learning
- regression model
- prior knowledge
- conditional probabilities
- neuro fuzzy
- hidden variables
- image sequences
- image segmentation