What is wrong with my matrix derivation?

Question

Below is my derivation, and result is different from formula $(15)$. What is wrong with my matrix derivation?

Answer 1

They use that $M$ and $\Lambda$ are symmetric, so it becomes

$$\frac{\partial \epsilon}{\partial Q} = \Lambda(Q-X) + \Lambda^T(Q-X) + \mu MQ + \mu M^TQ = 2\Lambda(Q-X) + 2\mu MQ$$

and when requiring

$$\frac{\partial \epsilon}{\partial Q} = 0$$

for maximization, you can cancel the $2$s:

$$\Lambda(Q-X) + \mu MQ = 0$$

Answer 2

Your derivation is correct. Conclude your result using the fact that $\Lambda$ and $M$ are symmetric matrices.