Lecture on common mistakes in ML

[nipunbatra]

• scaling train/test differently
• optimise on test set
• etc.

[patel-zeel]

We can include some content from bnomial and also from our experience with common gotchas of ML.