Online Learning
A training paradigm where the model updates continuously as new data arrives, one example at a time (or in small batches), rather than training on a fixed dataset.
Why It Matters
Online learning enables real-time model adaptation for applications like ad click prediction and stock trading where data distributions change constantly.
Example
A recommendation engine that updates its predictions after each user interaction — learning from every click and purchase in real time.
Think of it like...
Like a restaurant that adjusts its menu daily based on what sold well yesterday, rather than changing it only once a year.
Related Terms
Continual Learning
Training a model on new data or tasks over time without forgetting previously learned knowledge. Also called lifelong learning or incremental learning.
Stochastic Gradient Descent
A variant of gradient descent that updates model parameters using a single random training example (or small batch) at each step instead of the entire dataset. It is faster and can escape local minima.