– Sponsoring shown for too long
– Structure is not good. The video is called “tabular data”, but mostly discusses other things. I would propose to move the part about supervised/unsupervised and classification/regression to the previous video (together with system, features, …). The part about tabular data already ends quite soon. And then you have a final part which I would combine with the part about “Exploratory data analysis” (next video).
– Giving the example of a tensile curve for tabular data is a bad idea. Because that is a vector for 1 data point. Better to take the example like for clustering: processing and properties of a single material (sample).
The explanation of what supervised / unsupervised is, should be improved. It is well explained in the summary, but not in the main part of the presentation. You don’t even mention that you need labelled data for supervised models.