Which of the following about the gradient descent is incorrect?

Question

Options

• ARandom gradient descent is a commonly used one in gradient descent.
• BGradient descent includes random gradient descent and batch gradient descent.
• CThe gradient descent algorithm 1s fast and reliable [Right Answers}
• DRandom gradient descent is one of the commonly used optimization algorithms in deep learning

• A
  15% (9)
• B
  8% (5)
• C
  73% (44)
• D
  3% (2)

Explanation

Option C is marked correct (as the incorrect statement) because gradient descent is neither inherently fast nor reliably guaranteed to converge to a good solution - it can be slow on large datasets, sensitive to learning rate choice, and prone to getting stuck in local minima or saddle points.

Why the other options are actually true (and therefore not the answer):

• A is correct: stochastic (random) gradient descent (SGD) is indeed a widely used variant within the gradient descent family.
• B is correct: gradient descent broadly encompasses both stochastic (random) and batch variants as its two primary forms.
• D is correct: SGD and its derivatives (Adam, RMSprop, etc.) are among the most dominant optimizers in deep learning practice.

Memory tip: Think of gradient descent like hiking downhill in fog - you can find a valley, but not necessarily the deepest one, and the path can be slow or erratic depending on your step size. "Fast and reliable" describes an ideal that gradient descent often doesn't meet, especially on complex loss surfaces.

No community discussion yet for this question.