1 Correction to: Comp. Appl. Math. (2017) 36:877–883 https://doi.org/10.1007/s40314-015-0266-8

In the proof of Theorem 3 in Salkuyeh et al. (2017a), to have

$$\begin{aligned} (\tilde{L} (\alpha ^{(1)}) - \tilde{L} (\alpha ^{(2)}))_{ij} = \sum _{k=1,k\ne i}^{n}(\alpha _{ik}^{(1)}-\alpha ^{(2)}_{ik}) a_{ik}a_{kj}\le 0, \end{aligned}$$

the assumption

$$\begin{aligned} \sum _{k=1,k\ne i}^{n}(\alpha _{ik}^{(1)}-\alpha ^{(2)}_{ik}) a_{ik}a_{kj}\le 0, \end{aligned}$$
(1)

should be added to the statement of theorem and its proof remains true without any changes.

In fact, the proof of Theorem 3 in Salkuyeh et al. (2017a) follows from a similar strategy used in Salkuyeh et al. (2017b, Theorem 2.3) where a restriction analogous to (1) had been already assumed in the statement of theorem; see Salkuyeh et al. (2017b, cf. (2.2)). Basically, the assumption (1) could be naturally set. However, after extension of results in the final revision, the condition (1) was removed by mistake from the statement of theorem.