Proper datasets #42
Description
@Saurabh7
Currently, all toy datasets are either non-sense, or not informative. I suggest we re-place them with meaningful stuff.
Like a blob of two Gaussians for classification toy. And a real classification dataset, you can come up with something nice.
For regression, the sine wave is cool, what about a real dataset like CO2