From the course: Introduction to Auditing AI Systems

Unlock the full course today

Join today to access over 24,500 courses taught by industry experts.

AI performance

AI performance

- [Instructor] In order to perform a good AI audit it's important to know the different types of ML models. What factors can impact model performance and common metrics used to measure performance. There are four main types of machine learning, supervised unsupervised, semi-supervised, and reinforcement learning. In supervised learning a computer is trained on a labeled dataset meaning that the desired outcomes are already known. A supervised model then makes predictions about what class a new example is in. Unsupervised learning, on the other hand involves training a model on an unlabeled dataset meaning that the desired outputs aren't known in advance. An unsupervised model then identifies patterns within the data on its own which often group similar data points together. Semi-supervised learning uses partially labeled and partially unlabeled datasets and it's often useful when labeling data can be expensive or time…

Contents