Please enable JavaScript.
Coggle requires JavaScript to display documents.
Chapter 20: Communicate Model Insights (types of information to be…
Chapter 20: Communicate Model Insights
types of information to be communicated
business problem
model quality metrics (confusion matrix)
areas where a model struggles
most predictive features for model building
feature types especially interesting to management
recommend business actions
Unlocking Holdout
make sure there haven't been any mistakes in the model creation process
a small set of data can sometimes lead to the wrong direction
look at ranking differences between cross validation and holdout scores
Business Problem First
the problem should've continually guided the work completed
consult a subject matter expert
outline the confusion matrix
Areas where the Model Struggles
two types of data
internal
predictive ability could be extracted
external
purchased data
Most Predictive Features
directional results make intuitive sense
Not All Features are Created Equally
4 kinds of features to consider
Features that need to be changed and therefore require a re-run of the models
features requiring further examination
Immutable features
mutable features