Treasure Data's primary idea portal.
Submit your ideas & feature requests directly to our product requirements team! We look forward to hearing from you.
Radom Forest (RF) is easy to use because it does not require intensive feature engineering when compared to other algorithms. RF is also preferred by many users because it is easy to understand predictions.
Wrote an tutorial describing how to use RandomForest classification on Treasure Data.
More tutorials would follows.
This is already implemented in Hivemall v0.4. We're waiting for the deployment. Stay tuned. > https://github.com/myui/hivemall/wiki/Iris-multi-class-classification-using-RandomForest
You won't be notified about changes to this idea.