Limitations of the hyperplane separation technique for bounding the extension complexity of polytopes.
Matthias BruggerPublished in: Oper. Res. Lett. (2021)
Keyphrases
- hyperplane
- convex hull
- support vector
- support vector machine
- feature space
- data points
- linear classifiers
- training samples
- worst case
- principal components
- kernel function
- incremental learning algorithm
- input space
- maximal margin
- linearly separable
- upper bound
- classification procedure
- linear separability
- svm classifier
- support vectors
- normal vectors
- locality sensitive
- pattern recognition
- input data
- supervised learning
- nearest neighbor
- average distance
- dynamic programming
- lower bound
- neural network