Please enable JavaScript.
Coggle requires JavaScript to display documents.
Provost Ch 6 (Data mining: right sort of data (retrive, classifications…
Provost Ch 6
Data mining: right sort of data
retrive, classifications regression, cluster, recommendation
goal: determine the value of target characteristics
measurement first step: geometry Euclidean distance
most similar instances: nearest neighborhoods
ex: whiskey
Combining function
Ex credit card
Probability estimation is more valuable than yes/no
n=3
1 more item...
Heterogeneous Attributes
Ex: Credit card
Dictionary of Distances
Manhattan distance
Whiskey used Jaccard Distance: Set of characteristics
Cosine Distance: text classification similarity of 2 documents
edit distance: minimal number required
similarity-moderated scoring equation
prespecified target characteristics
1 more item...