In a neural network, knowing the weight and deviations of each neuron is the most important step. If you know the exact weights and deviations of neurons in some way, you can approximate any…

Question

Options

• ARandom assignment, pray that they are correct
• BSearch for a combmat1on of weight and deviation until the best value 1s obtained
• CAssign an initial value to iteratively update weight by checking the difference between the best
• DThe above is not correct

• A
  13% (7)
• B
  7% (4)
• C
  77% (43)
• D
  4% (2)

Explanation

Option C describes gradient descent with backpropagation - the foundational training algorithm for neural networks - where weights and biases (deviations) are initialized to small random values, then iteratively adjusted by measuring the error (loss) between the network's output and the true answer, and nudging each parameter in the direction that reduces that error. Option A (random assignment with no updates) is wrong because luck alone cannot reliably find good weights in a space with millions of dimensions. Option B describes an exhaustive combinatorial search, which is computationally intractable - the weight space is continuous and astronomically large, making brute-force enumeration impossible in practice. Option D is incorrect because option C accurately captures the real training process.

Memory tip: Think of the word "iteratively" in option C as your anchor - neural network training is never a one-shot guess (A) or an exhaustive search (B), but a loop of predict → measure error → adjust, repeating until the network converges on good weights.

No community discussion yet for this question.