11. Learning: Identification Trees, Disorder

Share it with your friends Like

Thanks! Share it with your friends!


MIT 6.034 Artificial Intelligence, Fall 2010
View the complete course: http://ocw.mit.edu/6-034F10
Instructor: Patrick Winston

In this lecture, we build an identification tree based on yes/no tests. We start by arranging the tree based on tests that result in homogeneous subsets. For larger datasets, this is generalized by measuring the disorder of subsets.

License: Creative Commons BY-NC-SA
More information at http://ocw.mit.edu/terms
More courses at http://ocw.mit.edu


Hammi says:

when a professor stars talking about vampries …
ok you got my atention

Simon S says:

On the one hand, that video provides a good fundament(and intuition) for understanding of decision trees.
On the other hand, I was really surprised how basic the material for TOP1 technical university in the world.

David Parry says:

Great video series. Thank you.

Zhenhao Zhou says:

Why no laptops?

Zhenhao Zhou says:

Why no laptops?

Andrei Chelaru says:

Fabulous teacher ! Regards from Romania !

Haniel Rameshbabu says:

A professor with a lot of energy.

Mohamed Benaichouche says:

Why does he make the difference between Identification tree and decision tree?

q zorn says:

i have a vampire in my tv set on friday fright night.

Bobby Nazaris says:

the blackboard buttons are making me crazy! I can not understand the sequences that he uses to move them up and down. Would someone tell me how they work please?! Thank you 🙂

Jabberwocky says:

After the "shadow" test, the disorder values for the other tests changed and the change was different for all tests.
Now it's not the case in this dataset, but is it possible that in larger datasets there are optimal test-trees that don't begin with the best test or even completely omit it? And if so, what would be the best general approach for building those test-trees?

James Liu says:

30:51: We use this because it's a convenient mechanism, it seems to make sense…", what a great teacher.

fgfgdfgfgf says:

So watered course 🙁 …

Snorkeldykkeren says:

Still, its not the correct format to program an artificial intelligence, this is about programming a software program. Not artificial intelligence or lifeforms. Artificial intelligences has to create these things by themselves…Learn them or "build" them

TheSharkasmCrew says:

very underrated prof imo

Marian Lenehan says:

Sorry, I still don't understand how you arrived at 4 in the shadow group. Is it because there were 4 question marks? I fully understand the other homogenous sets. Thanks.

kyoungd says:

I don't want to be a blood sucking vampire, so thanks professor.  Identification trees and disorder.  Now we are talking.  🙂

Write a comment


Area 51