ML

In the Ultra Transcenders books

AZ-305AI-901AZ-900DP-750SC-200DP-600DP-700

Each book explains ML in context, with comparison tables and the common traps.

Terms in this definition

• PREDICT

  A Fabric function for scoring MLflow models in batches from a notebook, provided the models have signatures; you can call it via MLFlowTransformer, Spark SQL or a PySpark UDF. Fabric Warehouse does not support the T-SQL PREDICT statement.

• Regression

  Predictive, rather than generative, supervised learning technique that outputs a number, for example rentals or revenue.

• Classification

  Predictive, supervised learning that assigns an item to one of a set of categories, for instance deciding whether an email is spam; it isn't generative.

• LIKE

  Compares strings with a pattern that can contain the % and _ wildcards. Because it only understands character patterns, searching big volumes of text this way is much slower than using full-text search.

• Azure Machine Learning

  Azure platform where you train, deploy, monitor and retrain models of your own, practising MLOps through tools such as pipelines, online endpoints, model monitoring and prompt flow.

• Generative AI

  AI that produces new content such as code, text, audio or images in response to a prompt. Predicting a class or a numeric value is predictive machine learning rather than generative AI.

Related terms

• AI

  Short for artificial intelligence. Language models and image recognition are examples: services allowing software to perceive, reason and create content, which Azure provides in the cloud next to machine learning.

• AI engineer

  A data job focused on adding AI capabilities to apps and data processes, using large language models, machine learning pipelines and data sources. Microsoft Learn associates this role with Microsoft Foundry.

• Apache Spark pool

  Spark compute in Azure Synapse used to build and update Delta Lake tables, run machine learning and stream changes from the Cosmos DB change feed.

• Automatic aggregations

  For DirectQuery semantic models, this feature studies the query log with machine learning and then creates and maintains in-memory summaries by itself; any aggregations you've defined manually continue to work next to them.

• Defender Queue Assistant

  A feature of the incident queue that rates every incident between 0 and 100 using machine learning and says what drove the score. Analysts can then tackle the most urgent incidents before the rest.

• Event Types

  Machine learning in VCF Operations for logs sorts look-alike messages into broad groups, each listed with an example and a rough tally.

• Explain the increase

  From a chart point's Analyze menu (line, bar or column charts), machine learning ranks the categories whose share moved most since the prior point; its twin, Explain the decrease, covers falls. RLS, DirectQuery and Top N filters all block it.

• Intelligent view

  Uses machine learning in Application Map to pick out probable reasons for failures and hide everything else; works for periods of up to seven days.