From the course: Automated ML.NET Training, Metrics, and Accuracy
Join today to access over 24,700 courses taught by industry experts.
Model metrics - ML.NET Tutorial
From the course: Automated ML.NET Training, Metrics, and Accuracy
Model metrics
- [Narrator] In this chapter, we'll be talking about model metrics. We'll also look at a demo for binary classification, and then examine permutation feature importance, or PFI, as well as doing a demo for that as well. Model metrics are specific to a model's machine learning task. For example, for binary classification, we're going to have different metrics than those for multi-class classification or regression or any of the other tasks. Likewise, the names of those metrics, depending on the task, will vary and the values that they provide will vary. Typically, values will range between zero and one, however that scale differs. In some cases, a value more towards zero is better, or in other cases, a value towards one is preferred. These metrics are available from the evaluate method, which is accessible from that specific task and the ML context object. Some of the metrics that are available for binary classification are accuracy, AUC, or area under curve, area under precision…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.
