Before you can use a machine learning (ML) model for match predictions, you must train and publish the ML model. Before you start training the ML model, you can review and add the match fields that you want to use in the training process. When you start the training process, the training data set is processed to generate record pairs. To train the ML model, label these record pairs that are presented to you in batches.
When you train the ML model, you can choose to label all the batches or conclude the training at the end of any batch. Labeling for a batch is complete after you label 30 record pairs with the conclusive labels Match and Not a Match. The training metrics, accuracy, precision, and recall, are calculated after a batch of record pairs is labeled. When you reach the desired level of accuracy, precision, and recall, complete the training process.
Note: The training process uses all the data that is available in the Business 360 data store.
1Open the business entity, and click the Match tab.
The Match Models page appears.
2Click the match model that was identified for training the ML model.
The Machine Learning Model page appears.
3In the Match Fields section, review the match fields.
The match fields that you configure for the declarative rules and the candidate selection criteria appear.
4To add additional match fields to train the ML model, click the Add Field icon.
The Add Field dialog box displays the fields apart from the existing match fields.
5After you review and add the match fields, click Start Training.
The training process starts preparing the data for training.
6After the training data preparation completes, click Start Labeling.
The Label Record Pairs dialog box appears.
7 Review and label record pairs in a batch.
- If you train the first ML model for your organization, review the process that the Label Record Pairs dialog box introduces, and click Start Labeling.
- If you train a subsequent ML model for your organization, record pairs are presented to you for labeling.
a Review and label a record pair with one of the following labels:
Label
Description
Not a Match
A conclusive label. Indicates that the records in the pair are different.
Need More Data
An inconclusive label. Indicates that you need more data to make a decision.
Not Sure
An inconclusive label. Indicates that you are not sure of your decision.
Match
A conclusive label. Indicates that the records in the pair are the same.
bTo review and label the next record pair, click Next.
c Continue labeling until you select conclusive labels for 30 record pairs in the batch.
After you select conclusive labels for 30 record pairs, the labeling for the batch is complete. To review the labeled record pairs, click Previous and Next.
8To complete labeling the batch, click Finish Batch.
The accuracy of the ML model is calculated.
9To label the next batch of record pairs, click Continue Labeling, and label record pairs by repeating step 7.
10 To resume labeling later, perform the following steps:
aTo pause labeling, click Close in the Label Record Pairs dialog box.
bTo resume labeling, click Continue Labeling on the Match Training page, and label record pairs by repeating step 7.
11After you complete labeling the required number of batches and reach an acceptable accuracy, precision, and recall, click Finish Training.
The training process completes and saves the ML model as part of the match model.