EngineersofAI LogoCoursesBlogResearch LabAI LettersThe LabCode BankInteractive 3DKodrEarnestJobs
Skip to main content
Interactive 3D/Online Learning
Playback
Learning Rate
η = 0.050
Options
Cumulative Regret
Online SGD:0.00
Batch GD:0.00
Step:0 / 80
Online SGD updates after every single point - immediate adaptation. Batch GD waits for all data - better when distribution is stationary.
Enable Non-stationary to see the decision boundary shift mid-stream. Online SGD adapts; batch GD lags behind.

Online Learning - Interactive Visualization

Online learning processes data one example at a time, updating the model after each observation. This is essential for streaming data and adapting to distribution shifts. The regret measures how much worse online learning performs vs the best fixed model in hindsight. This visualization streams data points and compares online SGD vs batch GD, including a non-stationary option where the data distribution shifts mid-stream.

  • Watch online SGD update model after each point vs batch GD waiting
  • Toggle non-stationary data to see online learning adapt to distribution shift
  • Watch cumulative regret accumulate over time
  • Adjust learning rate to see stability-plasticity tradeoff
  • Foundation for recommendation systems, financial modeling, and robotics

Part of the EngineersOfAI Interactive 3D - free interactive visualizations covering every major concept in machine learning and AI engineering. Hover any element for a plain-English explanation. No code required.