There are some notational confusions: at 4:32, you use Error (h), then write this P(CXORh) and then use "Pr". What is difference between Error (h) and P(CXORh)? For probability you use both "P" and "Pr", that is not consistent.
1/2 is just a conveniently small value that can be made arbitrarily small. Having error over 50% or probability less than 50% would make for a useless model
There is big difference in reading and teaching
After some time I think I've understood it thanks to this video. Very clear explanation!!
Never saw a great explanation than this ❤️
Beautifully explained. Thank you!
Great explanation. Thank you
Detailed explanation and good PPT. I like it!
There are some notational confusions: at 4:32, you use Error (h), then write this P(CXORh) and then use "Pr". What is difference between Error (h) and P(CXORh)? For probability you use both "P" and "Pr", that is not consistent.
Great explanation. Thanks!
Great video!, Thanks for explanation
Very nice
Thank You maam, really well explained
Thank you very much that was really helpful
Amazing!!!
Can I get ppt of this video?
Asalu aaa error values ela vesaru?epsilon =0.05 endhuku
14:40 where does inequality eq came from
By applying Hoeffinding Inequality
thanks
Ma'am could please share us the notes if possible.
thank you. but can u give me its coding?
We can read ,,,,
Why epsilon and delta is upper bounded by 1/2?
1/2 is just a conveniently small value that can be made arbitrarily small. Having error over 50% or probability less than 50% would make for a useless model