Azure Cookbook

Chapter 10. Azure Cognitive Services

Machine learning (ML) is a subset of artificial intelligence (AI). It’s the process of developing and training mathematical models that use data to help computers learn without direct instructions. The ML model is used to improve the performance and accuracy of a task.

Microsoft Azure employs AI/ML to enrich its service capabilities. AI-driven adaptive protection enables Microsoft Defender to adapt to new types of attacks and further protect your services. Azure Firewall threat intelligence-based filtering denies traffic from malicious IP addresses and URLs. Azure SQL Advanced Threat Protection uses AI/ML (and other sources) to detect anomalies indicating possible attempts to exploit databases. Microsoft leverages AI/ML not only to enhance Azure services but also to enable Azure clients to use AL/ML to solve their own business problems.

Creating a trained model can be time consuming, needs big training datasets, and is generally performed by field experts. Data scientists can use services such as Azure Machine Learning or Azure Databricks to build, deploy, and manage ML models. These services are recommended for AI/ML specialists who need to train their own new models. They generally take the following steps to create, optimize, and employ machine learning models for a given problem/scenario:

1. Choose the right machine learning algorithm based on the problem at hand.

2. Prepare training datasets by processing raw data.

3. Use the training datasets ...