Enhanced Kalman gain in RLS algorithm for identification accuracy with low excitation and heavy noise.
Fabien CourregesJolan ChassagnollePublished in: Int. J. Model. Identif. Control. (2014)
Keyphrases
- high accuracy
- computational complexity
- computational efficiency
- computational cost
- k means
- objective function
- learning algorithm
- dynamic programming
- identification rate
- probabilistic model
- worst case
- filtering algorithm
- kalman filter
- detection algorithm
- neural network
- preprocessing
- expectation maximization
- np hard
- gaussian noise
- feed forward neural networks
- noise immunity
- classification accuracy
- markov random field
- input data
- cost function
- optimal solution
- accuracy rate
- similarity measure
- bias variance