A matrix-free algorithm based on Newton's method for overdetermined nonlinear system of equations.
Ariel Sanhueza R.Claudio E. TorresPublished in: SCCC (2017)
Keyphrases
- high accuracy
- computational cost
- improved algorithm
- dynamic programming
- detection algorithm
- cost function
- preprocessing
- clustering method
- detection method
- significant improvement
- theoretical analysis
- computationally efficient
- mathematical model
- experimental evaluation
- recognition algorithm
- objective function
- optimization method
- k means
- decomposition method
- matching algorithm
- learning algorithm
- tree structure
- coefficient matrix
- qr decomposition
- similarity measure
- computational complexity
- probabilistic model
- optimization algorithm
- newton raphson
- segmentation method
- energy function
- input data
- segmentation algorithm
- support vector machine svm
- classification algorithm
- square root
- sparse matrix
- covariance matrix
- pairwise
- optimal solution
- distance matrix
- support vector machine
- linear systems
- matrix representation
- feature selection
- algebraic equations
- transformation matrix
- levenberg marquardt
- similarity matrix
- expectation maximization
- evolutionary algorithm
- principal component analysis