Step 7: Item-Clustering - Run Clustering, View Trade-Off, Choose Cluster Count

Step 7.1: Cluster Items

This API call executes the clustering using the latent factors/embeddings trained in the recommender. You select the minimum and the maximum number of item-clusters to be tested for their MSEs (Mean Square Errors). This will be used to choose the optimal number of item-clusters for your organisation.

For more information, please refer to the API documentation.

Step 7.2: Trade-Off Item Chart

This API call gets the mean squared error for each item-cluster count between the specified min and max.

The API will return an MSE for each cluster count. We can then use this data to decide how many clusters we want to set in the upcoming API call.

For more information, please refer to the API documentation.

Step 7.3: Choosing Item Cluster Count based on MSE Results

Once you have looked at the clusters' MSE results, you can decide how many item clusters are optimal. Usually, the number of clusters where the drop-off in MSE elbows is selected.

For more information, please refer to the API documentation.