Classifying more nonfraudulent records as fraudulent records


Problem: A data mining routine has been applied to a transaction dataset and has classified 88 records as fraudulent (30 correctly so) and 952 as nonfraudulent (920 correctly so).

a) Lowering the cutoff below 0.5, leads to classifying more nonfraudulent records as fraudulent records (more 0's are misclassified as 1's)

b) Lowering the cutoff below 0.5 leads to classifying more fraudulent records as nonfraudulent records (more 1's are misclassified as 0's)

Request for Solution File

Ask an Expert for Answer!!
Other Subject: Classifying more nonfraudulent records as fraudulent records
Reference No:- TGS03311042

Expected delivery within 24 Hours