Machine Learning

Problem Types

Pick the right setup before you pick the model.

Common ML Problems

  • Classification: predict a category (spam/not spam).
  • Regression: predict a number (price, demand).
  • Clustering: group similar items without labels.
  • Ranking/Recsys: order items (search, recommendations).
  • Time series: forecast with ordered data (often needs time-aware splits).

Fast Decision Rule

# If your target is:
# - category -> classification
# - number   -> regression
# - none     -> clustering / anomaly detection