Statements (53)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:model
|
| gptkbp:abbreviation |
gptkb:SVM
|
| gptkbp:advantage |
effective in high dimensional spaces
memory efficient robust to overfitting (with proper regularization) works with non-linear data (using kernel trick) |
| gptkbp:basedOn |
learning theory
|
| gptkbp:canBe |
linear kernel
polynomial kernel radial basis function kernel sigmoid kernel |
| gptkbp:hasApplication |
bioinformatics
diagnosis financial forecasting image recognition handwriting recognition face detection text classification spam detection |
| gptkbp:hasConcept |
hyperplane
maximizing margin support vectors |
| https://www.w3.org/2000/01/rdf-schema#label |
Support Vector Machines
|
| gptkbp:implementedIn |
gptkb:TensorFlow
gptkb:LIBSVM gptkb:Weka gptkb:scikit-learn R |
| gptkbp:introduced |
gptkb:Alexey_Chervonenkis
gptkb:Vladimir_Vapnik |
| gptkbp:introducedIn |
1990s
|
| gptkbp:limitation |
difficult to interpret
not suitable for very large datasets requires careful parameter tuning sensitive to feature scaling |
| gptkbp:relatedTo |
gptkb:machine_learning
gptkb:Lagrange_multipliers binary classification pattern recognition quadratic programming C parameter hard margin kernel trick large margin classifier multiclass classification nonlinear classification soft margin structural risk minimization |
| gptkbp:usedFor |
gptkb:dictionary
regression outlier detection |
| gptkbp:bfsParent |
gptkb:Journal_of_Machine_Learning_Research
|
| gptkbp:bfsLayer |
5
|