can statedifference betweenexclusive,overlapping, andfuzzy clusteringfinds amistake inthe slidedidn't checkemail duringentirepresentationdidn't usephone duringentirepresentationBrought thereadingmaterialsto classAnswered aquestion thepresenteraskedCORRECTLYcan draw thedifferencebetweenhierarchical &partitionalclusteringcan givea reasonfor usingclusteringanalysiscan describeand drawthreedifferent typesof clustersknows how tochoose thenumber ofclusters in K-meansclusteringCan namethree basicclusteringalgorithmsReadingmaterialshaveannotationsand highlightsknowsdifferencebetweencomplete andpartialclusteringSaid helloto onlinestudentsAsked aquestion thatconfusedeverybodycan describethe ideabehind K-meansclusteringAsked aquestion thepresenterdidn't knowthe answer toparticipatedin at leastonediscussionAnswered aquestion thepresenteraskedINCORRECTLYclapped atthe end ofpresentationAsked afollow upquestionto aquestionGave ausefulexample toexplain aconceptdidn'tfallasleep!Asked aquestion thepresenterdidn't knowthe answer tocan statedifference betweenexclusive,overlapping, andfuzzy clusteringfinds amistake inthe slidedidn't checkemail duringentirepresentationdidn't usephone duringentirepresentationBrought thereadingmaterialsto classAnswered aquestion thepresenteraskedCORRECTLYcan draw thedifferencebetweenhierarchical &partitionalclusteringcan givea reasonfor usingclusteringanalysiscan describeand drawthreedifferent typesof clustersknows how tochoose thenumber ofclusters in K-meansclusteringCan namethree basicclusteringalgorithmsReadingmaterialshaveannotationsand highlightsknowsdifferencebetweencomplete andpartialclusteringSaid helloto onlinestudentsAsked aquestion thatconfusedeverybodycan describethe ideabehind K-meansclusteringAsked aquestion thepresenterdidn't knowthe answer toparticipatedin at leastonediscussionAnswered aquestion thepresenteraskedINCORRECTLYclapped atthe end ofpresentationAsked afollow upquestionto aquestionGave ausefulexample toexplain aconceptdidn'tfallasleep!Asked aquestion thepresenterdidn't knowthe answer to

Clustering and Community Detection - Call List

(Print) Use this randomly generated list as your call list when playing the game. There is no need to say the BINGO column name. Place some kind of mark (like an X, a checkmark, a dot, tally mark, etc) on each cell as you announce it, to keep track. You can also cut out each item, place them in a bag and pull words from the bag.


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
  1. can state difference between exclusive, overlapping, and fuzzy clustering
  2. finds a mistake in the slide
  3. didn't check email during entire presentation
  4. didn't use phone during entire presentation
  5. Brought the reading materials to class
  6. Answered a question the presenter asked CORRECTLY
  7. can draw the difference between hierarchical & partitional clustering
  8. can give a reason for using clustering analysis
  9. can describe and draw three different types of clusters
  10. knows how to choose the number of clusters in K-means clustering
  11. Can name three basic clustering algorithms
  12. Reading materials have annotations and highlights
  13. knows difference between complete and partial clustering
  14. Said hello to online students
  15. Asked a question that confused everybody
  16. can describe the idea behind K-means clustering
  17. Asked a question the presenter didn't know the answer to
  18. participated in at least one discussion
  19. Answered a question the presenter asked INCORRECTLY
  20. clapped at the end of presentation
  21. Asked a follow up question to a question
  22. Gave a useful example to explain a concept
  23. didn't fall asleep!
  24. Asked a question the presenter didn't know the answer to