Foundation for the New Algorithm Learning Pseudo-Independent Models.
Jae-Hyuck LeePublished in: ECSQARU (2005)
Keyphrases
- learning algorithm
- experimental evaluation
- learned models
- times faster
- fully connected
- incremental learning
- learning speed
- clustering method
- search space
- computational complexity
- dynamic programming
- preprocessing
- computational cost
- probabilistic model
- worst case
- learning process
- learning models
- objective function
- bayesian framework
- inductive inference
- optimal solution
- bayesian methods
- parameter estimation
- learning phase
- training data
- similarity measure
- cost function
- active learning
- prior knowledge
- k means
- learning rules
- neural network
- learning scheme
- parametric models
- hidden variables
- learning tasks
- optimization algorithm
- expectation maximization
- linear programming
- high accuracy
- np hard
- significant improvement