Substantial improvements in the set-covering projection classifier CHIRP (composite hypercubes on iterated random projections).
Leland WilkinsonAnushka AnandDang Tuan NhonPublished in: ACM Trans. Knowl. Discov. Data (2012)
Keyphrases
- set covering
- random projections
- set covering problem
- dimensionality reduction
- dimension reduction
- integer programming
- original data
- random sampling
- principal component analysis
- fourier transform
- feature space
- image reconstruction
- column generation
- training data
- sparse representation
- hash functions
- feature selection
- low dimensional
- high dimensionality
- frequency domain
- feature set
- support vector machine
- training set
- data sets
- document clustering
- active learning
- combinatorial auctions
- input data
- nearest neighbor
- support vector
- multiscale
- integer program
- computer vision
- learning algorithm