Answer-first summary for fast verification
Answer: Collect a set of labeled training documents, train the model using Azure Custom Document Understanding, evaluate its performance, and publish it when ready.
To implement a custom document intelligence model, you should first collect a set of labeled training documents that represent the specific type of document you want to extract data from. Then, you can use Azure Custom Document Understanding to train the model on these documents. After training, you should evaluate the model's performance using a separate set of test documents. If the model meets your accuracy requirements, you can publish it for use in your Document Intelligence solution. Option B is incorrect because manually correcting errors is not a scalable solution for implementing a custom model. Option C is incorrect because creating a custom Azure AI Search skill is not the appropriate approach for training a custom document intelligence model. Option D is incorrect because Azure Machine Learning is not the recommended tool for building custom document intelligence models.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
No comments yet.
Your team is considering implementing a custom document intelligence model to extract data from a specific type of document that is not well-supported by the prebuilt models. What steps should you follow to train, test, and publish this custom model?
A
Collect a set of labeled training documents, train the model using Azure Custom Document Understanding, evaluate its performance, and publish it when ready.
B
Use the prebuilt models to extract data from the documents and manually correct any errors.
C
Create a custom Azure AI Search skill and use it to extract data from the documents.
D
Use Azure Machine Learning to build and train a custom machine learning model.