how simple and helping! It turned out that doing it practically & interactively makes it more understandable instead of just writing long equations. Thank You Sir!!
its the total number of distinct words that are occurring in the given documents. so those six are Chinese, Beijing, Shanghai, Macao, Tokyo, Japan. Rest are the repetition of those words.
Yes I agree. I watched this video 2 years ago. When I needed today, I searched whole youtube for this specific video. This is such a nice example with a really nice explanation. Edited: punctuation
It's a very nice video, thank you! If I may give a bit of advice, imho, if "word bag" is defined and "features" of each document is explicitly written, it might be easier to understand for new learners. Great video though, thanks!
Dear Rafael Merino García! Thank for your presentation. I have a problem with Multinomial Naive Bayes. I can't fully understand the meaning one of the fragment in the formula of the probability of a document in Multinomial Naive Bayes Model. P(di|Cj) = P(|di|). |di|!. U(P(Wt|Cj)^Nit / Nit !) with i = 1, .., |V|. U is Integration, comment isn't allowed for special symbol so I can't express it. My question: P(|di|), what does this probability mean? How to compute it? Please explain for me! Thanks you so much. Best regard! Nam.
For future purpose : If the text that we are trying to classify has not been occurred even once then its probability would be 0. However, this would make the entire sentences probability to be 0. To avoid this we add 1 which he adds in the video as well that there is some probability of it being in any of the category. Adding 1 is part of laplacian smoothing.
I'm not sure, just an idea: Calculate conditional probabilities of the words for the third, fourth....nth class. Then find P(c3|d5), P(c4|d5), ..... P(cn|d5) using the same approach. The P(ci|d5) with max value will be the most probable class of the sample d5.
Seriously, this is a good example which easily to understand.
how simple and helping! It turned out that doing it practically & interactively makes it more understandable instead of just writing long equations. Thank You Sir!!
really simple, finally I understand multinomial naive bayes
Such an awesome example - Really helpful for understanding with an example👍👍
could you please explain how you got the size of vocabulary |V| as 6?
its the total number of distinct words that are occurring in the given documents. so those six are Chinese, Beijing, Shanghai, Macao, Tokyo, Japan. Rest are the repetition of those words.
Thank you. You know how to make things obvious!
Yes I agree. I watched this video 2 years ago. When I needed today, I searched whole youtube for this specific video. This is such a nice example with a really nice explanation. Edited: punctuation
whats the different multinomial naive bayes and the basic naive bayes?
this video really helped me...thanks!!
Dear Rafael Merino García, thank you!!
Wow, really good tutorial. Best wishes from Poland
very good explanation, very very good, thank you for posting this
It's a very nice video, thank you! If I may give a bit of advice, imho, if "word bag" is defined and "features" of each document is explicitly written, it might be easier to understand for new learners. Great video though, thanks!
Why do we have to
smoothing for Chinese, they both
appear in 2 classes
Dear Rafael Merino García! Thank for your presentation. I have a problem with Multinomial Naive Bayes. I can't fully understand the meaning one of the fragment in the formula of the probability of a document in Multinomial Naive Bayes Model.
P(di|Cj) = P(|di|). |di|!. U(P(Wt|Cj)^Nit / Nit !) with i = 1, .., |V|. U is Integration, comment isn't allowed for special symbol so I can't express it.
My question:
P(|di|), what does this probability mean? How to compute it?
Please explain for me! Thanks you so much.
Best regard!
Nam.
Perfect explanation
how to compute P(d5)?
Why did you not include the word 'Sanghai' ? Or did I not hear you ignoring it intentionally since I watched at 2x speed?
Why you listen by a speed of 2x? hhhhh
where is the multinomial?
What if there are more than one feature?
Hi, I am one of them ;)
Thanks for this. So helpful.
Thank you man...........
What if there is unknown word in testing document?
Probability = 0
For future purpose :
If the text that we are trying to classify has not been occurred even once then its probability would be 0.
However, this would make the entire sentences probability to be 0.
To avoid this we add 1 which he adds in the video as well that there is some probability of it being in any of the category. Adding 1 is part of laplacian smoothing.
Thank you, this was really helpful!
thank you sooo much.
thanks
very good
I don't see any multinomiel here.
same here
I saw naive but not multinomial
Thanks!
to the point
how about giving credit to the author of the example and the slides? (Dan Jurafski) You are explaining everything as if it was your work.
how to compute multi-label classification using this classifier ?
I'm not sure, just an idea: Calculate conditional probabilities of the words for the third, fourth....nth class. Then find P(c3|d5), P(c4|d5), ..... P(cn|d5) using the same approach. The P(ci|d5) with max value will be the most probable class of the sample d5.