Question: Which preprocessing step is crucial when using Stochastic Gradient Descent ( SGD ) to ensure convergence towards the global optimum? Question options: Stratified sampling of
Which preprocessing step is crucial when using Stochastic Gradient Descent SGD to ensure convergence towards the global optimum?
Question options:
Stratified sampling of training data
Dimensionality reduction of feature space
Normalization of training instances
The training instances are shuffled during training
Step by Step Solution
There are 3 Steps involved in it
1 Expert Approved Answer
Step: 1 Unlock
Question Has Been Solved by an Expert!
Get step-by-step solutions from verified subject matter experts
Step: 2 Unlock
Step: 3 Unlock
