Intermediate · 30 minYou are hereRead
Gradient descent
Make a loss small by stepping against its gradient, and see the learning rate decide whether the steps crawl, glide, zig-zag or fly off.
Foundations
Linear regression and gradient descent
Gradient descent makes a loss small by repeated steps against its gradient. This topic fits linear models with many features, by least squares and by descent, speeds the steps up with small random batches, and draws curves with transformed features.
Start with Gradient descent0 of 4 read
Lessons on this topic
Related topics