Icon

02.1 Batch predictions via scheduling

<p>This workflow demonstrates how to calculate predictions in <strong>batch</strong> on a regular basis. It can be <strong>scheduled</strong> on KNIME Hub.</p>

02.1 Batch predictions via scheduling

This workflow demonstrates how to calculate predictions in batch on a regular basis. It can be scheduled on KNIME Hub.

Dates filtering
Select data forpredictions, e.g., basedon execution date
Spark SQL Query
Select featuresfor predictions
Spark Column Filter
Add a timestamp
Spark SQL Query
Connect to yourDatabricks workspace
Databricks Workspace Connector
Spark Missing Value (Apply)
Apply
Spark Predictor (Regression)
Space Connector
Space Connector
Execution date
Date&Time Configuration
KNIME Hub Authenticator
Save predictionsand append new
Spark to Parquet
Preprocessing modelfor missing values
Model Reader
Import Unity datato a Spark DataFrame
Parquet to Spark
Connect and start yourall-purpose computecluster on Databricks
Create Databricks Environment
TrainedML model
Model Reader

Nodes

Extensions

Links