Icon

H2O AutoML Regression Model

H2O AutoML for Regression

This example shows how to build a regression model with H2O AutoML, predict new data and score the regression metrics for model evaluation.

1. Prepare:
Load the Combined Cycle Power Plant data, import the resulting KNIME Table to H2O and partition the data for test and train set 20/80.

2. Learn:
We learn the regression model using the "H2O AutoML Learner (Regression)" node using the default algorithm settings but limiting the max runtime to 60 seconds. The node learns and optimizes multiple different models and selects automatically the best one. The node outputs are the selected model and a leaderboard showing the metrics for the different models. The first row of the leaderboard is the selected model.

3. Predict:
Make predictions on test data using the selected model.

4. Score:
In order to evaluate our model, we assess different metrics by scoring the predictions made on the test data.

Import data to H2O FrameSplit betweentraining and test setStart local H2O NodePredict ontest setScore predictionLoad Combined Cycle Power Plant dataApply AutoML on training setTable to H2O H2O Partitioning H2O Local Context H2O Predictor(Regression) H2O Numeric Scorer Excel Reader H2O AutoML Learner(Regression) Import data to H2O FrameSplit betweentraining and test setStart local H2O NodePredict ontest setScore predictionLoad Combined Cycle Power Plant dataApply AutoML on training setTable to H2O H2O Partitioning H2O Local Context H2O Predictor(Regression) H2O Numeric Scorer Excel Reader H2O AutoML Learner(Regression)

Nodes

Extensions

Links

Download

To use this workflow in KNIME, download it from the below URL and open it in KNIME:

Download Workflow
Created by: nico
Created at: 2017-07-10
On NodePit since: 2021-08-06
Last update: 2024-05-01
Created with KNIME version: v4.4.0
Tags: H2Omachine learningautomlregressionautomatedautomatic

Deploy, schedule, execute, and monitor your KNIME workflows locally, in the cloud or on-premises – with our brand new NodePit Runner.

Try NodePit Runner!