Notebook improvements: 02_ml_pipeline.ipynb

[kate19brown]

1. More spacing in text
2. Avoid italics
3. More comments in code to explain code
4. list(falklands_df.columns) is the line necessary - as the output is very long
5. train/test split - could be clearer - test/train split 20% and 25% is confusing
6. Explain features even if they are small e.g. %%time
7. Consider if using sklearn.metrics.confusion_matrix() and sklearn.metrics.ConfusionMatrixDisplay() could make code simplier?