# Matching with machine learning from an algorithmic standpoint

Matching with machine learning involve using the following matching components:

* tMatchPairing
* tMatchModel
* tMatchPredict

To get an overview of the workflow used to predict matches for datasets automatically, see [Data Matching with Talend tools](https://help.qlik.com/talend/r/en-US/8.0/data-matching).

### In this section

* [How does tMatchPairing compute the suspect duplicate pairs?](https://help.qlik.com/talend/en-US/data-matching/8.0/how-does-tmatchpairing-compute-suspect-duplicate-pairs)
* [How does tMatchPairing compute the sample of suspect duplicate pairs?](https://help.qlik.com/talend/en-US/data-matching/8.0/how-does-tmatchpairing-compute-sample-of-suspect-duplicate-pairs)
* [How does tMatchModel learn a matching model?](https://help.qlik.com/talend/en-US/data-matching/8.0/how-does-tmatchmodel-learn-a-matching-model)
* [How does tMatchPredict predict values on a dataset?](https://help.qlik.com/talend/en-US/data-matching/8.0/how-does-tmatchpredict-predict-values-on-dataset)