What is online learning versus batch learning?

• A. Online learning updates with each new sample; batch learning trains on a fixed dataset in bulk.
• B. Online learning stores all data in memory; batch learning uses data streams.
• C. They are the same.
• D. Online learning only uses synthetic data.

Answer: (A)

The key idea is how the model learns over time. Online learning updates the model as each new data point arrives, allowing it to adapt quickly to changing patterns or streams of data. In contrast, batch learning trains the model on a fixed dataset all at once and uses that final model until the next full retraining. This leads to a stable solution when the data sample is representative, but it doesn’t adapt until you retrain.

That’s why the best choice says online learning updates with each new sample and batch learning trains on a fixed dataset in bulk. The other options aren’t accurate: online learning doesn’t require storing all data in memory and can work with streaming data, batch learning isn’t about data streams by default, and online learning isn’t limited to synthetic data.