Intrinsically Semi-supervised Learning

Description

These methods directly optimize an objective function with components for labeled and unlabeled samples and do not rely on any intermediate steps or supervised base learners. Basically, these methods are extension of existing supervised methods to include the effect of unlabeled data samples in the objective function.

Technical Details

Framework	MITRE D3FEND
Ontology URI	d3f:IntrinsicallySemi-supervisedLearning
Local Identifier	IntrinsicallySemi-supervisedLearning
Publication Status	Exists in ontology only

Relationships

Parent Tactics

D3A-SSL Semi-Supervised Learning
Model Model

Child Concepts

D3A-MML Maximum-margin Learning
D3A-PBL Perturbation-based Learning
D3A-SSML Semi-supervised Manifold Learning
D3A-SSGML Semi-supervised Generative Model Learning

Related Techniques

D3A-SSTL Semi-supervised Transductive Learning (Unpublished)
D3A-SSIL Semi-supervised Inductive Learning (Unpublished)
D3A-SSWM Semi-supervised Wrapper Method (Unpublished)
D3A-UP Unsupervised Preprocessing (Unpublished)
D3A-SSML Semi-supervised Manifold Learning (Unpublished)
D3A-GBSSL Graph-based Semi-supervised Learning (Unpublished)
D3A-SSL Semi-Supervised Learning (Unpublished)
D3A-SSGML Semi-supervised Generative Model Learning (Unpublished)

D3FEND