Hidden Markov Models 08: motivating the forward-backward algorithm

[@DataTales-AI]

Having this simple robot example is the best way to explain HMMs. Thank you so much !!

[@mohammedalmarakby1221]

exactly!

[@djp3]

Glad it was helpful!

[@antonokhotnikov]

Thank you so much Prof. Patterson, this is a beautiful explanation!

[@djp3]

Glad it was helpful!

[@justnazmus8140]

Did you study pattern analysis?

[@batcheneden6135]

THANK YOU!!!! by far the best HMM series on KZbin. thank you very much for this, you really cleared things up for me

[@djp3]

Glad it was helpful!

[@ananyakapoor2677]

Perfectly explained and extremely intuitive -- thank you!

[@jasmeetgujral5665]

Your videos are so underrated. They are really good.

[@djp3]

Glad you like them!

[@murtazaraja7510]

Excellent series. Thank you very much

[@homataha5626]

my problem is I can't code this material! I think I have understood the material but codingg them is so hard for me esp in a higher-dimensional space? can someone guide me on how to code them?

[@fgfanta]

The robot example seems contrived to me, also because I think, for any state, the robot can only produce one observation (symbol), the floor of the room where it is. Perhaps I haven't quite understood the example. The lesson is very clear nevertheless.

[@djp3]

Yes, it's absolutely contrived because that's not how real robots work.

[@xntumrfo9ivrnwf]

Your "B" is what other sources call the emission probabilities, right? Thanks

[@djp3]

yes.

[@sheikhakbar2067]

Very helpful series. Thanks.

[@djp3]

Glad it was helpful!

[@franciscoaristi613]

Excellent explanation!

[@djp3]

Glad it was helpful!

[@preetgandhi1233]

Great Explanation!

[@shreyassingh4032]

ryan reynolds teaches computer science !!!!!!

[@taggebagge]

Bless you.

[@endgamers7945]

hulk?