subpopulation analysis
omar21
Dataiku DSS Core Designer, Dataiku DSS Adv Designer, Registered Posts: 18 ✭✭✭✭
I'm trying to evaluate the performance of a model with the metrics and subpopulation analysis,
but I couldn't interpret the subpopulation analysis I don't understand the red/green flesh and the
numerical values associated to every attribute
can you help me
Answers
-
Manuel Alpha Tester, Dataiker Alumni, Dataiku DSS Core Designer, Dataiku DSS ML Practitioner, Dataiku DSS Core Concepts, Dataiku DSS Adv Designer, Registered Posts: 193 ✭✭✭✭✭✭✭
Hi Omar,
DSS splits the source dataset into two populations: training and test. The model is then evaluated against the test population.
The red numbers indicate the false-negatives (predicted 0 by the model, but actually 1) and false-positives (predicted 1, actually 0), in the test population. The green numbers represent correct predictions.
I hope this helps.