Please enable JavaScript.
Coggle requires JavaScript to display documents.
Evaluate Model Performance (Fraction of Variance FVE (This provides…
Evaluate Model Performance
All about maintaining the balance and momentum of the data
Fraction of Variance FVE
This provides binomial data
how much variance is in the dataset
it show how far off the data is
Decision tree classifier
Effective
Simple
Decide how the DTC will work with min complexity
One-=hot encoding is the move
ROC curve
Confusion matrix
optimizing metric
Prediction distribution
DataRobot determind the F1 score that maximizes the thrshold
Ture positive rate TPR
Later in the chapter discusses talking about
Comparing Model Pairs
Prioritizing Modeling Criteria & selecting model
Predictive accuracy.
Prediction speed.
Speed to build model. 4. Familiarity with model. 5. Insights.
Dual lift curve
ENET Blender
Using the ROC curve
Prefect model
Random model