Pathfinder's Identify Segments feature allows you to find groups within your data that are similar to each other based on the columns in your data.
Within Identify Segments, you can choose your preferred Segmentation Method:
- Clustering is best for grouping data points by numerical similarity.
- Network Extraction is best for grouping data points that have similarities across various categories.
Pathfinder will also review your dataset and recommend a Segmentation Method to use.
Using Clustering
To Identify Segments using Clustering:
- On the Pathfinder start screen, click Identify Segments.
- Choose Clustering.
- The ideal number of clusters is automatically chosen by default. To override this, enter the number of clusters manually in the input field.
- Click Identify Segments.
Pathfinder will populate with insights about key Segments, identified as clusters of data that are generated by running the Clustering AI routine.
Results
Target KPIs
Adding a Target KPI will adjust the insights to focus on how that particular KPI relates to the identified Clusters. To choose a KPI, navigate to Target KPI at the top of Pathfinder and click Select a Column. Click to choose a suggested column or any of the listed columns. You can clear any selected KPI by clicking the Remove icon.
Segment Details
You can Sort by Segment Size. Click the dropdown under Sort By to adjust the sorting.
For more details on each identified Cluster, click the Cluster to review:
- Common characteristics within the Cluster
- Statistical differences compared to other Clusters
- KPI Metric median value in the selected Cluster compared to the rest of the data
Key Drivers
Key Drivers shows the degree to which other columns in the data have impacted how the data was separated into Clusters. These drivers are automatically sorted by Most Impactful, and the level of impact is shown as a bar chart under each column name.
You can change the sorting to see the Least Impactful by clicking the dropdown under Sort By and selecting that option.
Using Network Extraction
To Identify Segments using Network Extraction:
-
- On the Pathfinder start screen, click Identify Segments.
- Choose Network Extraction.
- Add a Key Identifier. A key identifier should represent the things in your data that you would like to group together. Examples include user IDs, product IDs, serial numbers, or account numbers.
To add a Key Identifier, click Select a Column. Within the drop-down that appears, choose from either Suggested Columns, or select a column from the All Columns list. For datasets with a large number of columns, you can Search for a column in the text box above the list. Click a column to select it.
- Click Identify Segments.
Pathfinder will populate with insights about key Segments, which are automatically generated leveraging the Network Extractor and Explainable AI (XAI) AI routines. These Segments are also visualized in a generated Network Graph.
Initially, the three largest Segments are shown in Pathfinder. Click Show All Segments to view all identified Segments.
Results
Target KPIs
Adding a Target KPI will adjust the insights to focus on how that particular KPI is impacted by the chosen column. To choose a KPI, navigate to Target KPI at the top of Pathfinder and click Select a Column. Click to choose a suggested column or any of the columns listed below. You can clear any selected KPI by clicking the Remove icon.
Segment Details
You can Sort by Segment Size. Click the dropdown under Sort By to adjust the sorting.
For more details on each identified Segment, click the Segment to review:
- Common characteristics within the Segment
- Statistical differences compared to other Segments
- KPI Metric impacts on the given Segment as compared to the rest of the data
Next Steps
To continue your guided exploration using Pathfinder, select one of the Next Steps. Or, to reset your experience using Pathfinder, click the Menu icon on the top right corner of the panel and click Reset.
Previous Article |
Next Article |