ML 3.3Free videoLesson 3 of 7 in Module 3
Gradient Descent for Linear Regression
Derive the gradient of the squared-error loss, see why one update reads every row, and know what stochastic and mini-batch descent buy.
Watch the lesson: Gradient Descent for Linear Regression
In this lesson8 parts
- 01One update, done by hand
- 02Deriving the gradient
- 03What one update costs
- 04One row per update
- 05Mini-batches
- 06The library's version
- 07From scratch, then the library
- 08Recap, your challenge, and what comes next
Builds on
This lesson stands on its own. These go deeper into what it uses.
Key terms: Gradient Descent for Linear Regression
The words this lesson introduces, each in one line. The module’s glossary collects them all, and the AI glossary has every term across the courses.
- batch gradient descent
- Every update uses every training row.
- stochastic gradient descent
- Each update uses one training row, in shuffled order.
- epoch
- One pass through every training row.
- mini-batch
- A small group of training rows used for one update.
- batch size
- The number of rows in a mini-batch.
- closed form
- A formula that gives the minimum directly, without iterating.
Quiz: Gradient Descent for Linear Regression
Your first pick on each question is the one that counts. Getting one wrong here is how the lesson sticks.