Skip to content
← All PyTorch tracks

Build a Decision Boundary

PyTorch

A complete classifier project: train your model, inspect its learning curve, and export a runnable experiment.

A model you can see

From training loop to decision boundary

Build a linear classifier with a learned bias, watch its loss change, and inspect the regions it predicts. Export your code with a dataset and a separate holdout experiment.

  1. Compute predictions and a numerically stable loss.
  2. Train with gradient descent and record the learning curve.
  3. Inspect your result, then run and adapt the exported script.

Prerequisites: tensor shapes, matrix multiplication, sigmoid and gradients. The project is CPU-friendly; a GPU is optional.

Review binary classification Try the other framework →
0 / 1 solved
Start →
  1. 1. Not solved yet. Decision Boundary Lab