Gradient descent is an optimization algorithm used to train neural networks. Its objective is to modify the model parameters until finding one of the cost function’s local minimum ( very efficient weights and bias values).

Question

Complete the sentence :  Gradient descent is an optimization algorithm used to train neural networks. Its objective is to modify the model parameters until finding…

Accepted Answer

**One of** the cost function’s **local minimum** (<=> very efficient weights and bias values)

Answer

**The** cost function’s **overall minimum** ( ⇔  optimal weights and bias values)

Answer

**The** cost function’s **overall maximum** (<=>  optimal weights and bias values)

Answer

**One of** the cost function’s **local maximum** (<=> very efficient weights and bias values)

Question from the Deep learning - Fundamentals test

Gradient descent is an optimization algorithm used to train neural networks. Its objective is to modify the model parameters until finding one of the cost function’s local minimum (<=> very efficient weights and bias values).