Section 1 - Instruction

You've learned about the core mathematical engine of machine learning: elementwise operations, broadcasting, and reductions. Let's practice applying these fundamental calculations.

Engagement Message

Ready to compute?

Section 2 - Practice

Type

Multiple Choice

Practice Question

What is the result of the following broadcasting operation?

A. tensor([[11, 2], [3, 4]]) B. tensor([[11, 12], [13, 14]]) C. tensor([11, 12, 13, 14]) D. This will cause an error.

Suggested Answers

  • A
  • B - Correct
  • C
  • D
Section 3 - Practice

Type

Fill In The Blanks

Markdown With Blanks

Fill in the blank to compute the sum of each column in the tensor.

Suggested Answers

  • 0
  • 1
  • None
Section 4 - Practice

Type

Sort Into Boxes

Practice Question

Sort these operations into the correct category.

Labels

  • First Box Label: Elementwise Op
  • Second Box Label: Reduction Op

First Box Items

  • Addition (+)
  • Multiplication (*)

Second Box Items

  • torch.mean()
  • torch.max()
  • torch.sum()
Section 5 - Practice

Type

Swipe Left or Right

Practice Question

Does the scenario describe broadcasting? Swipe left for Yes, right for No.

Labels

  • Left Label: Broadcasting
  • Right Label: Not Broadcasting

Left Label Items

  • Adding a matrix (3,4) and a vector (4,)
  • Multiplying a matrix (2,2) and a scalar
  • Adding a matrix (3,4) and a vector (3,1)

Right Label Items

  • Adding two matrices of the same shape (2,2)
  • Multiplying a vector (5,) and another vector (5,)
Section 6 - Practice

Type

Multiple Choice

Practice Question

To normalize a feature column in a dataset (a 2D tensor), you first calculate the mean of that column. Which operation would you use to get the mean of the second column (index=1)?

A. torch.mean(data) B. torch.mean(data, dim=1) C. torch.mean(data[:, 1].float()) D. torch.mean(data[1, :])

Suggested Answers

  • A
  • B
  • C - Correct
  • D
Sign up
Join the 1M+ learners on CodeSignal
Be a part of our community of 1M+ users who develop and demonstrate their skills on CodeSignal