A Toolarge Weight Initialized On The Neural Network Can Lead To #524
A too-large weight initialized on the neural network can lead to vanishing gradient whereas a too-small weight initialized on the neural network can lead to exploding gradient.
This multiple choice question (MCQ) is related to the book/course
gs gs127 SQLite.
It can also be found in
gs gs127 Neural Networks in Machine Learning - Random Initialization - Quiz No.1.
