Registre d'evidència del mètode

Adversarial Training

Adversarial Training is a robust optimization procedure for deep neural networks in which the model is trained not on clean data alone but on worst-case perturbed inputs crafted during training. Formalized by Madry et al. (2018) as a min-max saddle-point problem, the method uses Projected Gradient Descent (PGD) to generate strong adversarial examples within a bounded Lp perturbation set before each gradient update, forcing the network to learn decision boundaries that are stable under such perturbations.

Sources recorded, not reviewed

Registre font

Les citacions es copien textualment del registre font del mètode. No s'infereix cap verificació a nivell de reclam d'elles.

Adversarial Training (Robust Optimization for DL)

Registre de mètode taxonòmic · ml-model / deep-learning

Madry, A., Makelov, A., Schmidt, L., Tsipras, D., & Vladu, A. (2018). Towards deep learning models resistant to adversarial attacks. International Conference on Learning Representations (ICLR). · URL

Obre el mètode complet

Reclamacions curades

Les reclamacions s'han persistit al registre de proves, cadascuna amb la seva pròpia avaluació.

Encara no hi ha reclamacions curades

Aquesta vista no inventa una avaluació de reclam quan el registre no en té cap.

Mètodes relacionats

Generat a partir del gràfic de mètodes i mostrat com a relacions suggerides per la màquina; no s'infereix cap reclamació d'evidència.

Taxonomic bucketData Augmentationmachine-suggested · Relational suggestion, not evidence.Same method familyGenerative Adversarial Networkmachine-suggested · Relational suggestion, not evidence.Same method familyOut-of-Distribution Detectionmachine-suggested · Relational suggestion, not evidence.