What is wrong description or backpropagation?

Question

Options

• AThe learning process or the backpropagation algorithm consists of a forward propagation process
• BThe backpropagation algorithm is a learning algorithm suitable for multi-layer neural networks,
• CThe backpropagation phase sends training inputs to the network to obtain an stimuli response
• Duntil the response of the network to the input reaches the predetermined target range.

• A
  6% (2)
• B
  9% (3)
• C
  82% (27)
• D
  3% (1)

Explanation

Option C is the wrong description because it actually describes forward propagation, not backpropagation. Sending training inputs through the network to obtain an output response is the forward pass - the backpropagation phase works in the opposite direction, using the computed error to calculate gradients and update the network's weights layer by layer.

• A is correct - backpropagation does include a forward propagation step as its first phase to produce an output before the error can be calculated.
• B is correct - backpropagation is specifically designed for multi-layer (deep) neural networks, where direct error attribution to each layer requires the chain rule.
• D is correct - the training loop (forward + backward passes) repeats until the network's output falls within a predetermined acceptable error range.

Memory tip: Think of backpropagation as a two-act play - Act 1 is the forward pass (input → output), and Act 2 is the backward pass (error → weight updates). If an exam option says the backpropagation phase sends inputs forward, that's the red flag - it's describing the wrong act.

No community discussion yet for this question.