What Is Online Learning? Updating a Model Continuously as New Data Arrives
Instead of training once on a fixed dataset, online learning updates a model continuously as new data streams in — with real benefits and real risks.
Updating continuously, rather than training once
Online learning updates a model incrementally as new data becomes available, one example (or small batch) at a time, on an ongoing basis — rather than training once on a large, fixed, already-collected dataset the way standard ("batch") training does.
Why this matters for constantly changing environments
In settings where the underlying patterns genuinely shift over time — trends, user behavior, real-time signals — a model trained once on historical data can gradually go stale. Online learning lets a model adapt continuously to new patterns as they emerge, without waiting for a full, separate retraining cycle.
The real risk that comes with continuous updating
A model updating continuously on live, real-world data can also drift in unintended directions if that incoming data is noisy, unrepresentative, or adversarially manipulated — online learning trades some of the stability and predictability of periodic, carefully controlled batch training for that ongoing adaptability.
Frequently Asked Questions
What is online learning?
A training approach that updates a model incrementally as new data becomes available, one example or small batch at a time, rather than training once on a large, fixed, already-collected dataset.
What is the main risk of online learning?
A model updating continuously on live data can drift in unintended directions if that incoming data is noisy, unrepresentative, or manipulated — trading some stability of periodic batch training for ongoing adaptability.