Choosing a Machine-Learning Task
Start with the output a user needs. “Use AI” is not a task definition; “route this message to one of 12 teams” is.
Translate the outcome into a task
| You need | Choose | Example output |
|---|---|---|
| A category from a spreadsheet row | Tabular classification | high_risk |
| A number from a spreadsheet row | Tabular regression | 42.7 MPa |
| A category from text | Text classification | card_arrival |
| Similar wording or likely duplicates | Lexical & fuzzy search | Ranked matching rows |
| Metrics over saved responses | Response evaluation | Completion and token F1 |
| One category for a whole image | Image classification | healthy_leaf |
| Objects and their locations | Object detection | Labelled bounding boxes |
| A class for every pixel or object mask | Segmentation | Pixel masks |
| Named landmarks | Keypoint detection | Joint coordinates |
Define the prediction moment
Write one sentence before configuring a model:
Using information available at this moment, predict this output for this kind of case, so that this user can take this action.
This sentence exposes leakage. If a churn model includes cancellation_date,
or a loan model includes a field created after approval, the held-out score can
look excellent while the deployed model is useless. Mark identifiers and
post-outcome fields as excluded features in Tabular AI.
Prefer the simplest useful baseline
A majority-class prediction, average-value regression, lexical search or small linear text classifier gives you a reference point. A more complex model earns its cost only when it improves the metric that matters while meeting latency, privacy and review requirements.
Know when the task needs redesign
- If labels are subjective, write annotation guidance and measure reviewer agreement.
- If rare cases matter, choose balanced accuracy, macro F1, recall or another metric that exposes them.
- If future data differs from historical data, use a time-aware test set outside the current random-split workflow.
- If the desired output is free-form generation, AnyLearning's current Text AI workflows can organize or evaluate saved text, but they are not an LLM trainer.