Model training

Train the algorithm using the k-Means node. The number of clusters (k) needs to be selected manually. Here, we set k=3.

Read data

The data contains various attributes about different houses and their price.

Pre-processing (data preparation)

Model evaluation

How to train a k-Means model?

Step 1: Drag the k-Means node and click on it to open the dialog.

Step 2: Set the "Number of clusters" to 3. In the "Column selection", include the columns "Lat" and "Long".

Step 3: Click on the node and "Execute" it to perform clustering.

How to evaluate a k-Means model?

Step 1:To visualize the clusters, add the ''Scatter Plot'' and the ''OSM Map View'' nodes. These nodes should be connected to the Color Manager to visualize clusters with colors.

Step 2:To evaluate the clustering task, connect the clustering output to the "Silhouette Coefficient" node. Select "Cluster" as the clustering column.

Step 3:Execute the node to get Silhouette Coefficients for each instance, each cluster, and for the overall clustering task.

Analyze Location Data by Clustering

In this workflow, we perform a clustering task on location data where we have longitude and latitude information. We use the k-Means algorithm to cluster this data and then visualize the clustering results.

Filter data

Keep only data for houses in California

Normalize data

Apply min-max normalization to latitude and longitude

Workflow complete!

Keep the momentum going by exploring Just KNIME It!on the Hub to challenge yourself and see how these nodes can be integrated into more complex workflows and use cases.