Hierarchical Clustering 3: single-link vs. complete-link

[bit.ly/s-link] Agglomerative clustering needs a mechanism for measuring the distance between two clusters, and we have many different ways of measuring such a distance. We explain the similarities and differences between single-link, complete-link, average-link, centroid method and Ward's method.

[@carlchen9935]

Thanks for the wonderful concise tutorials! But it would be much nicer if it is possible for you to draw stuff on the slides instead of somewhere not visible in the video (probably on the blackboard I guess). :)

[@YogeshDarji99]

Can you point your arrow, when you say distance between "that" and "that". Thanks, nice video!

[@INVFootSoldier]

Very good video, greetings from Poland University of Silesia, much appreciated, helped a lot

[@tetamusha]

Great lecture. It's hard to find examples of all these distance measures at work, specially in a single example and in such an educative way.

[@cosplayfx]

THANK YOU. I really like your exmplanations and graphics but as a none english man it's pretty hard to follow, though. understanding the topic AND language makes it hard but your graphical way to demonstrate the topic is great! keep going!

[@MBH-1997]

we dont see the board !! explain on the slide or show us the bored please ...

[@lamis_18]

unbelievable explanation I was wondering Why we should take the farthest clusters in COMPLETE LINKAGE ,, after this great explanation I understand now that only while counting d between all pairs of clusters but in any way I am gonna choose the minimum d whatever the linkage criteria is. THANK YOU VERYMUCH.

[@japalaia]

These are great instructional videos. Do you happen to have an instructional video on the BIRCH method. Specifically, I am trying to understand how the CF tree is constructed.

[@kenskyschulz1979]

I think one statement which is not very much vivid and i.e. cluster distance measure gives an intuition for combining two clusters. If not right then please reply.

[@kenskyschulz1979]

+Kensky Schulz Technically, distance measures are obtained to merge any two nearest clusters.

[@mg4649]

I liked this video pretty much. I would prefer to see what you were doing in the blackboard but I got the idea and confirmed my knowledge on Ward's algorithm. Thanks! :)

[@darthcaedus5124]

pictures are the best part of the video

[@touristvideoclips]

good!

[@vlavrenko]

Παντ Γιαννακοπουλος Thank you!

[@nz8960]

thank you. very helpful

[@AAFenta]

Very helpful!

[@frankzhang6009]

very clear

[@razanmasood4226]

many thanks!

[@arpitmathur2933]

Nice

[@computer-science-with-mani]

i specially came here for understanding what single linkage long chains actually means, and that was described somewhere in the background :( lol

[@kushagraraj8798]

It is wrong. Your concepts are not clear.

[@MamSirProductions]

Thanks for the effort, but if i can give you one tip: try more examples and less talking... I didn't find what I was looking for in this video in the end, BUMMERRRR

[@vlavrenko]

Sorry to hear that. Perhaps the answer was in one of the other videos for this lecture. The full lecture is here: th-cam.com/video/GVz6Y8r5AkY/w-d-xo.html

+Victor Lavrenko well the replier is right. you don't give examples but just formulas like zillions of others. practical examples what community need to learn and understand

+Victor Lavrenko can you make a video about wards method with example? ty

[@jumongxoxo]

andrew tate?

[@srummel]

Thank you for your videos - they are very clear and concise and present the complete ideas in logical, comprehensible 'chunks'. Your work is helping me with my graduate work immensely. I am very grateful.