Comparison of the Convergence Rates of the New Correntropy-Based Levenberg-Marquardt (CLM) Method and the Fixed-Point Maximum Correntropy (FP-MCC) Algorithm.
Ahmad Reza HeraviGhosheh Abed HodtaniPublished in: Circuits Syst. Signal Process. (2018)
Keyphrases
- convergence rate
- levenberg marquardt
- conjugate gradient
- least squares
- fixed point
- initial guess
- artificial neural networks
- dynamic programming
- back propagation
- objective function
- step size
- similarity measure
- training algorithm
- bundle adjustment
- global convergence
- convergence speed
- learning algorithm
- primal dual
- learning rate
- dynamical systems
- particle swarm optimization
- input data
- linear programming
- probabilistic model
- optimal solution
- closed form
- policy iteration
- pairwise
- approximation methods
- neural network