Add more data processing steps to the classification tutorial

[lars-reimann]

The current classification tutorial is missing important steps:

• Remove some columns completely
• Impute age and maybe fare
• Reverse transform the resulting table.

It should also use headings instead of an ordered list to populate the table of contents. Ideally, it should match the "Getting started" tutorial for the DSL closely, so it's easy to compare DSL and Python syntax.

[LIEeOoNn]

with "reverse transform the resulting table" is meant to just reverse the OneHotEncoding, because we imputed missing values with SImpleImputer which isn't invertible

[lars-reimann]

with "reverse transform the resulting table" is meant to just reverse the OneHotEncoding, because we imputed missing values with SImpleImputer which isn't invertible

Correct.