Techniques & Methods
Online Learning
Online learning (also called incremental or streaming learning) updates model parameters one example or small batch at a time as data arrives, enabling the model to adapt to changing patterns without full retraining. It is used in recommender systems, fraud detection, and financial forecasting.
Online learning presents challenges including catastrophic forgetting (overwriting old knowledge with new) and sensitivity to noisy data. Techniques like elastic weight consolidation help mitigate forgetting.
Authority Links
Related Terms
Techniques & Methods
Training
Teaching a model to make accurate predictions by exposing it to large datasets.
Miscellaneous
Training Data
The labeled or unlabeled dataset used to fit a model's parameters during the learning process.
Techniques & Methods
Reinforcement Learning
An agent learns by taking actions in an environment and receiving rewards or penalties.
Techniques & Methods
Fine-Tuning
Continuing the training of a pre-trained foundation model on a smaller, curated dataset to specialize its behavior, style, or domain expertise without losing its general capabilities.