Please enable JavaScript.
Coggle requires JavaScript to display documents.
Chapter 18: Comparing Model Pairs (Model Comparison (To understand and…
Chapter 18: Comparing Model Pairs
Model Comparison
To understand and examine the differences between the overall best model and the best non-blender model
Contains each algorithm's predictions by probability from high to low and splits them into bins
Helps select the appropriate model for implementation and which to proceed with
Hospital readmission example
Prioritizing Modeling Criteria and Selecting a Model
5 Criteria
Predictive accuracy
Accuracy comes at a cost
Prediction speed
How quickly you need results based on your time constraints
Speed to build model
How long it takes to train a model
Depends on how much data the model is trained with and the complexity of an algorithm
Familiarity with model
Able to understand the exact meaning of its results
Insights
Based on the assumption that different algorithms make different statistics available (ex: regression results)
Should know the different measures generated through different prediction distribution thresholds and their impact on the confusion matrix
After going through the steps, one model should stand out