Please enable JavaScript.
Coggle requires JavaScript to display documents.
Chapter 16. Understanding the Process (16.1 Learning Curves and Speed…
Chapter 16. Understanding the Process
16.1 Learning Curves and Speed
Important to see if it will improve predictive ability
Diminishing Marginal Improvement: Greater the amount of relevant data
Leaderboard tab
Learning Curves
Shows validation scores for X and Y axis
Y axis: lower score is more favorable
Dots represent
Regularized Logistic Regression
1 more item...
16.2 Accuracy Tradeoffs
Speed vs Accuracy tab
how rapidly the model will evaluate
efficient fronteir line
speed needs to be evaluated against peak time
always look for efficient fronteir line
16.3 Blueprints
Blueprint pane
shows that different models exist
converts into different categorical features
16.3.1 Imputation
highest ranked
Regularized Logisitc
uses median for value feature
16.3.2 Standardization
Standardize box
Different Deviations
"scaled"
One-Hot Encoding
16.3.3 One-hot Encoding
Shows if original feature has only two values
True or False
Split categorically
Newborn
parameters used to simplify data
Open algorithm box
16.4 Hyperparameter Optimization
Advanced tuning