Edit content on data pre-processing and limitations

[agitter]

We currently discuss data pre-processing in the T-cells lesson and give examples there. We do not teach specific data cleaning strategies because they are domain specific. However, we can do more to stress the importance of data cleaning so that participants are not mislead. The choice of classifier is not relevant if the data are of poor quality.

In addition, we can close the workshop with some discussion of limitations. These includes limitations of ML in general and limitations of participants' knowledge after the workshop. That can transition into the presentation of next steps and resources that participants can use to continue their ML education.