Hello,
I am done with this course. Honnestly, I am learning ML by my own and for my own interest (ok final objective is to predict lotofoot results:))
I have to admit this course is just great. Of course I will still need to deep dive some concepts but your learning approach is giving possibilities for experimentations. Time consuming but very usefull. Globally I found the wrap up tests interesting.
Maybe one aspect could be added in this course : I understand it is difficult to have focus on all type of modelling methods. Nevertheless I feel that as introduction could be usefull :
-
To give overview of all model families and their main focus points (for example SVM models, when to use, for what purpose, advantages & drawbacks…
-
Give some high level workflow examples : for beginners the different method details like test/train spit vs Cross validation vs Gridsearch vs different CV methods can be at the end quite confusing. Maybe as an introduction giving some generic workflow would help to have the big picture needed to design model.
But thanks to this course I believe at the end I understand now better what I am doing… for example using attrition datatset from kaddle, got following results (still far to be perfect !!)
Thnaks a lot for patience, have nice vacations…
