Bayesian Network - Exact Inference Example (With Numbers, FULL Walk-Through)

Рет қаралды 36,426

John McVickar

Күн бұрын

Timestamps
Relevant Equations - 0:12
Brief Aside - 1:52
Example Problem - 2:35
Solution - 3:41

Пікірлер: 25

@Fakevomet 2 жыл бұрын

This video is saving my life right now

@mhrumi8994 4 ай бұрын

Thanks a lot, I was looking for this all over the Internet for weeks. ❤

@chiaral8516 3 жыл бұрын

Thank you! This is very helpful having all the exact calculations and also going over all the relevant formulas up front is super. The calculations of the probabilities across the probability tables went a bit fast for me but I'll break it down and study it tomorrow. This is the only full example I found after scouring the internet, looking in tesctbooks, and reviewing the lecture videos for tge course Im in.

@alishafarman4370 3 жыл бұрын

kzfaq.info/get/bejne/o5qcabhi27qye40.html

@harshsason3725 7 ай бұрын

you are a life savor. Thanks

@mohamedalaqel4570 3 жыл бұрын

Thanks! this is very useful

@mattanova 2 жыл бұрын

FINALLY! Thank you

@CowboyRocksteady Жыл бұрын

excellent video on bayes nets. thank you

@prachimehta9839 2 жыл бұрын

Amazing Video !

@janenie51 Жыл бұрын

Thank you so much🎉🎉🎉

@CamelEnjoyer 4 жыл бұрын

Wow co video John, very cool!

@alishafarman4370 3 жыл бұрын

kzfaq.info/get/bejne/o5qcabhi27qye40.html

@tianlongwang7238 Жыл бұрын

Hi there, I like your matrix solution so much that make the computation so easy, I also calculate the same question in Naive bayesian Classifier, thich give me an astonishingly close result, which is 0.28461. I want to investigate more on the reasons that easier architecture can achieve a similar accuracy as a more sophisticated architecture, do you or anyone who saw this comments have any ideas?

@OpeLeke 2 жыл бұрын

This video is loaded. do you have another one where you go into details?

@reallyidrathernot.134 2 жыл бұрын

mates! What's the capital letters instead of lowercase letters mean? eg: P(a) turning into P(A)?

@dfreeze371 Жыл бұрын

For anyone still confused by this: P(A) is the total probability distribution of A, which can then turn into two distinct outcomes P(a) and P( ¬ a)

@niharamzan382 2 жыл бұрын

why have you ignored alpha in your computation?

@adityachawla7523 2 жыл бұрын

you can just normalize in the end to account for alpha

@sakibahmed4317 Жыл бұрын

why don't you have other videos

@bmyalbara2 3 жыл бұрын

But how to define the values of probabilities used in this example or in any other example?

@froylanm.wbariomartinez470 3 жыл бұрын

Excuse me, do you know how to define it now ? thanks

@bmyalbara2 3 жыл бұрын

@@froylanm.wbariomartinez470 It is either you have historical data that you can extract the probabilities from by dividing the number of times a specific event happens over the total number of events, or, in case you do not have data, you can ask experts in a specific field to provide their assessment of the probabilities. Read about (Expert Judgement).

@trouvaille5965 2 жыл бұрын

@@bmyalbara2 Thank you so much