PCA Checkpoint

1. PCA finds directions of:

Maximum variance Minimum variance Zero variance Constant values

2. Principal components are:

Orthogonal to each other Parallel to each other Random directions Always positive

3. Before PCA, data should typically be:

Centered (mean-subtracted) Sorted Shuffled One-hot encoded

4. PCA uses which matrix decomposition?

5. Reducing dimensions with PCA helps with:

Visualization and noise reduction Increasing model complexity Adding more features Perfect accuracy