Gavin (36) [Avatar] Offline
#1
You show a formula in a diagram - but unless you have a degree in Mathematics / Statistics / Category Theory - it doesn't really mean much to me.

In fact this entire section is really quite hard work to digest; Especially in the introduction chapter.

Explaining the initial vocabulary and the differences between graph types directed / non-directed / etc is quite appropriate here in the introduction - but everything else seems too theoretical for the very first chapter of the book.

There just seems to be too much in an introduction.

I;m not saying it isn't important data. But I am not seeing the value of it here - especially without concrete examples that I can wrap my head around.

Even in your equation, right here, a completed example using one of the previous diagrams would help prove how the math makes sense - and why it is important to me, the reader.

Perhaps a sentence or two after the formula about how the average degree in a graph aids / proves complexity / assists in choosing an appropriate algorithm...

In the intro you don't need to entertain what a high or low value means / dictates / leads to - just that it is a useful piece of information to be known, in the overall ML project?
Alessandro Negro (9) [Avatar] Offline
#2
which formula are you referring to?
Anyway the first chapter introduce some basic concepts and terminology so that people can understand the rest of the book.
I tried to keep them as simple as possible and minimalistic by reducing them to what is really important.

Do you have some concrete examples that are hard to understand? I'll try to make them more clear in the next version.
601063 (3) [Avatar] Offline
#3
Hi,
I see no formula that would not be understandable by everyone with a bot of willingness to search by himself.

Maybe you are referring to the sigma symbol ? If yes just look it up, you can see it like a loop that performs a sum on what's inside.

I just want to say that that I absolutely do no want to see this book emptied out of its mathematical content, which many books intended for programmer tend end up doing, dramatically reducing their quality and density.

A serious book on an advanced topic cannot possibly be entirely self-sufficient, and that is, in my humble opinion, perfectly fine.
Sure, some might need to look a few things up; but not re-explaining things from the very beginning saves a lot of time to the writer and spares dozens boring paragraphs to the advanced reader.
Also, once you've read a few books, it becomes quite annoying to see each and everyone of them re-explain the exact same things just a bit differently, in order to earn the "No prerequisite" flag (which is now almost a red one for me).
Alessandro Negro (9) [Avatar] Offline
#4
Trust me, mathematic will not disappear. I totally agree with you. Math concepts are part of science and specifically data science.