# Feature importance report

When using the tMatchModel component in a Job, you can generate a report to see the feature importance in the model.

The feature importance report contains:

* The heat map of all features,
* The summary of the settings of the tMatchModel component and
* The model quality.

The heat map helps you quickly see the importance of:

* Each feature,
* Each matching key and
* Each metric.

The summary displays the model quality and is a reminder of the parameters you set to generate this model.

For more information on how to generate this report, see the [tMatchModel properties for Apache Spark Batch](https://help.qlik.com/talend/en-US/data-matching/8.0/tmatchmodel#tmatchmodel-properties-for-apache-spark-batch) of the tMatchModel component.

### In this section

* [Presentation of the feature importance report](https://help.qlik.com/talend/en-US/components/8.0/matching-with-machine-learning/presentation-of-feature-importance-report)
* [Analyzing the heat map](https://help.qlik.com/talend/en-US/components/8.0/matching-with-machine-learning/analyzing-heat-map)
* [Improving a matching model](https://help.qlik.com/talend/en-US/components/8.0/matching-with-machine-learning/improving-matching-model)