Combining Physics-Based Domain Knowledge and Machine Learning using Variational Gaussian Processes with Explicit Linear Prior.
Daniel L. MarinoMilos ManicPublished in: CoRR (2019)
Keyphrases
- gaussian processes
- domain knowledge
- machine learning
- bayesian methods
- dirichlet process
- gaussian process
- prior knowledge
- preference learning
- gaussian process regression
- covariance function
- multi task learning
- model selection
- hyperparameters
- human pose estimation
- gaussian process models
- multi task
- decision trees
- machine learning algorithms
- nonlinear regression
- computer vision
- learning algorithm
- prior information
- closed form
- support vector machine
- active learning
- data mining
- learning problems
- bayesian inference
- image reconstruction
- non stationary
- pairwise
- training data