Parallelization of the nearest-neighbour search and the cross-validation error evaluation for the kernel weighted k-nn algorithm applied to large data dets in matlab.
Ginés RubioAlberto GuillénHéctor PomaresIgnacio RojasBen PaechterPeter GlösekötterC. I. Torres-CeballosPublished in: HPCS (2009)
Keyphrases
- nearest neighbour
- knn
- cross validation
- k nearest neighbour
- k nearest neighbor
- data sets
- k nearest
- training set
- nearest neighbor
- error estimates
- support vector
- k means
- learning algorithm
- data analysis
- neural network
- unseen data
- training data
- prior information
- expectation maximization
- euclidean distance
- em algorithm
- generalization error
- rough sets
- mixed data
- kernel function
- high dimensional data
- similarity search
- distance function
- text categorization
- hyperparameters
- artificial neural networks
- machine learning
- data points