A Novel Pre-Processing Method to Reduce Noise Effects in a Prototype-Based Clustering Algorithm.
Zahra TaghikhakiBehrouz MinaeiAlireza MasoumPublished in: IKE (2008)
Keyphrases
- preprocessing
- computational cost
- hybrid method
- synthetic data
- clustering method
- high accuracy
- clustering algorithm
- high precision
- filtering method
- objective function
- additive noise
- pairwise
- dynamic programming
- noisy data
- similarity measure
- probabilistic model
- classification method
- neural network
- noise removal
- noise immunity
- noise reduction
- detection method
- support vector machine
- experimental evaluation
- cost function
- significant improvement
- computational complexity
- multiscale
- face recognition
- image processing