Please enable JavaScript.
Coggle requires JavaScript to display documents.
Ch. 20 Communicate Model Insights (Info that should be communicated during…
Ch. 20 Communicate Model Insights
Info that should be communicated during a presentation
Business Problem
Model quality metrics (confusion matrix)
Areas where a model struggles (potential for improvement through more
data––features & cases)
Most predictive features for model building
Feature types especially interesting to management (e.g., insights into the
business problem and unknowns uncovered during the modeling process)
Recommended business actions (i.e., to implement model or not, any
business decisions to implement at various probability thresholds, and how will doing so change practice?)
Unlocking Holdout
ensure no mistakes were made in the model creation process
done by releasing the holdout data - comparing cross-validation with holdout scores
1) order of models did not change = best outcome
2) at least top models stay at top of list = 2nd best outcome
Business Problem First
The problem should have also been refined as the AutoML process carried forward, given that the analytics made clear more detailed information about the model and features of the data
Areas where model struggles
two main types of data
internal data
external data
Pre-processing and Model Quality Metrics
model metrics from the confusion matrix were combined to better understand model performance characteristics
procuring data, cleaning that data, and carefully addressing issues
Most predictive features
directional results tend to make intuitive sense
Not All Features are Created Equally
Features that need to be changed and therefore require a re-run of the models
Features requiring further examination
Immutable features
Mutable features
Recommended Business Actions
explicit recommendations for next steps
three concrete recommendations are available:
1)Implement the model to find the 62% of patients most likely to be readmitted
2) Institute a one-month pilot program targeting five percent of patients discharged to home
3) Institute a data-extraction-and-purchase pilot program to explore how the patient readmission model can be improved and to what degree.
James Frainey
jafr4672@colorado.edu