Glossario IA
Il dizionario completo dell'Intelligenza Artificiale
Support Vector Machine (SVM)
Supervised learning method that constructs a hyperplane or a set of hyperplanes in a high-dimensional space to separate different data classes.
Hyperplane
Subspace of dimension one less than the ambient space, used in SVMs as a decision boundary to separate data classes.
Margin
Distance between the decision hyperplane and the closest data points (support vectors), maximized to improve the robustness of the SVM model.
RBF Kernel (Radial Basis Function)
Commonly used kernel function in SVMs that maps data into an infinite-dimensional space, based on the Euclidean distance between points.
Nu (ν) Parameter
Parameter of the One-Class SVM that controls the expected proportion of anomalies in the training data and influences the model's tolerance to errors.
Novelty Detection
Variant of anomaly detection where the model is trained on normal data and then used to identify new classes not seen during training.