5 Comments
User's avatar
David Sutherland's avatar

Thanks for writing posts set to this size. I like having bite sized posts for me to digest.

Expand full comment
Tivadar Danka's avatar

Thanks for the feedback! I'm currently experimenting with shortest posts, and I'm glad that you like it!

Expand full comment
Sani Nassif's avatar

Something is odd about the plots pre/post the application of the reciprocal. The yellow lines are not such that 1 / first = second.

Expand full comment
Tivadar Danka's avatar

I'll share the generating code soon!

Expand full comment
jmzc's avatar

Thanks.

In this example, why is it necessary to transform the linear function in logistic regression? The function ax + b can already make good predictions, why do we need a non-linear transformation?

Expand full comment