Question

Suppose we have a linearly separable dataset, and we divide the data into training and validation sets. Will a perceptron learned on the training dataset (assuming gradient decent works perfectly well) be guaranteed to have i) 0 error on the training dataset ii) 0 error on the validation dataset. Brie y explain. (10 points)

Answer 1

Answer:

Theoretically Yes

Explanation:

The data given is linearly separable. So, the subset of the data will also be linearly separable. And it will pass for all training dataset. However, you should definitely never expect such thing In any real-life problem because the data is noisy, for a bazilion of reasons, so no model is guaranteed to perform perfectly.