On the Minimax Optimality of the EM Algorithm for Learning Two-Component Mixed Linear Regression.
Jeongyeol KwonNhat HoConstantine CaramanisPublished in: AISTATS (2021)
Keyphrases
- em algorithm
- linear regression
- maximum likelihood
- expectation maximization
- hidden variables
- least squares
- mixture model
- parameter estimation
- learning algorithm
- linear predictors
- locally weighted
- log likelihood
- parameter learning
- likelihood maximization
- likelihood function
- maximum likelihood estimation
- gaussian mixture
- unsupervised learning
- principal component analysis
- supervised learning
- video sequences
- reinforcement learning