Explanation

Option A is correct because the weight update in neural networks follows the gradient descent formula:

$w_{new} = w_{old} - \eta \cdot \frac{\partial E}{\partial w}$

Where:

• $w_{old} = 0.20$
• $\eta = 0.10$ (learning rate)
• $\frac{\partial E}{\partial w} = 0.25$ (partial derivative)

Calculation: $w_{new} = 0.20 - (0.10 \times 0.25) = 0.20 - 0.025 = 0.175$

Option B (0.195) would result from adding the gradient instead of subtracting it.

Option C (0.225) would result from incorrectly multiplying the learning rate by the weight instead of the gradient.