Machine Learning

Train classification or regression from gold datasets, or try Housing / Iris samples.

Before you start

Path	Need
Your data	Gold published for ML, status ready, 100+ rows
Samples	Nothing. Use Sample datasets in the wizard

Sidebar → Data Intelligence → Machine Learning (/ml)

Compared to Forecast and Loop

	ML	Forecast	Loop
Output	Category or number per row	Future values over time	Predictions that learn from feedback
Data	Gold (or samples)	Forecast gold (or AirPassengers)	Loop wizard (not Lake gold)

Full chooser: What is Heimdall?

Path

1. Lake → gold or samples
2. Build a model
3. Deploy
4. Monitor

How training works: How training works.

Common mistakes

Mistake	Fix
CSV upload in the wizard	Only samples. Use Lake → gold
Empty gold picker	Publish with Machine learning use case, status ready

Next

• Build a model
• ML API integration