Machine Learning Fundamentals · free preview
Programming with Data
Learning replaces hand-written rules with a parameterized model fitted to examples.
1. Writing rules by hand
Writing rules by hand
2. The rule we can follow
The rule we can follow
3. What must be in place before training
What must be in place before training
4. Why this framing matters
Why this framing matters
5. What the learning algorithm produces
In the wake-word example, the dataset holds audio snippets and two-way labels. The model is a fixed architecture whose parameters are adjusted against the dataset. Which statement describes what the learning algorithm produces?
6. The three things to settle before training
Name the three items a designer must fix before a learning algorithm can run. Give them in the order this lesson presents them: ____, ____, and ____.
Practice this interactively
Free account · instant grading · spaced review that schedules itself.
Start this course for free