BLOCK 3 - SAMPLING & RESAMPLIN
MACHINE
CLASS
LEARNING
💡 TOPICS COVERED: Monte Carlo Integration, Cross-Validation, and Boot
Recall Notes
Cross-Validation Statistical technique
(Validation-Set
Involve partitioning the dataset into two sm
approach)
Training set - typically 60% to 80% of th
Validation set - 10% to 20%
Perform multiple cross-validations to reduce
Advantages
Conceptually simple and easy to implement
single model fit and evaluation
Disadvantages
Highly dependent on how the data is split
validation sets, which can result in high
estimated test error
Since fewer observations are available fo
splitting the dataset, the model may tend
test error rate
What is K-Fold Cross- Iterative process instead of performing vali
Validation?
Randomly splits the dataset into K number of
For a total of K iterations, one fold will b
model, while the remaining K-1 is used for t
MACHINE
CLASS
LEARNING
💡 TOPICS COVERED: Monte Carlo Integration, Cross-Validation, and Boot
Recall Notes
Cross-Validation Statistical technique
(Validation-Set
Involve partitioning the dataset into two sm
approach)
Training set - typically 60% to 80% of th
Validation set - 10% to 20%
Perform multiple cross-validations to reduce
Advantages
Conceptually simple and easy to implement
single model fit and evaluation
Disadvantages
Highly dependent on how the data is split
validation sets, which can result in high
estimated test error
Since fewer observations are available fo
splitting the dataset, the model may tend
test error rate
What is K-Fold Cross- Iterative process instead of performing vali
Validation?
Randomly splits the dataset into K number of
For a total of K iterations, one fold will b
model, while the remaining K-1 is used for t
