Learning to Adapt: A Method for Automatic Tuning of Algorithm Parameters.
Jamie SherrahPublished in: ACIVS (1) (2010)
Keyphrases
- learning algorithm
- preprocessing
- cost function
- high accuracy
- fine tuning
- detection algorithm
- optimization algorithm
- parameter tuning
- expectation maximization
- similarity measure
- computational complexity
- theoretical analysis
- parameter settings
- improved algorithm
- em algorithm
- input data
- significant improvement
- detection method
- learning scheme
- objective function
- parameter selection
- clustering method
- experimental evaluation
- dynamic programming
- computationally efficient
- initial values
- computational cost
- recognition algorithm
- matching algorithm
- optimal parameters
- parameter space
- segmentation method
- k means
- synthetic and real images
- prior knowledge
- levenberg marquardt
- parameters estimation
- probabilistic model
- estimation algorithm
- energy function
- convergence rate
- reinforcement learning
- fully automatic
- classification algorithm
- segmentation algorithm
- density function
- adaptation process
- support vector machine svm
- nonlinear functions
- tree structure
- bayesian methods
- distance metric
- pairwise
- optimal solution
- optimization method
- learned models
- initial estimates