An Improved Penalty Algorithm using Model Order Reduction for MIPDECO problems with partial observations.
Dominik GarmatterMargherita PorcelliFrancesco RinaldiMartin StollPublished in: CoRR (2021)
Keyphrases
- objective function
- probabilistic model
- mathematical model
- computational complexity
- learning algorithm
- dynamic programming
- energy function
- parameter estimation
- em algorithm
- partial observations
- orders of magnitude
- expectation maximization
- linear programming
- probability distribution
- optimal solution
- domain specific
- semi supervised
- state space
- state variables
- approximate solutions
- machine learning