We show that Chen et al.'s schemes [IEEE TCC, 2(4), 2014, 499-508] for outsourcing linear regression computation to the cloud are not unquestioned. In scheme 1, the client has to generate an orthogonal matrix. Its computational complexity is almost equal to that of solving a linear regression problem locally. In such case, the client has no necessary to outsource the computations to the cloud. In scheme 2, it masks a matrix by multiplying two diagonal matrixes. The linear transformation is very vulnerable to statistical attack.

We remark that the Wang et al.'s scheme [IEEE Transanction Parallel Distributed Systems, 24 (6), 1172-1181, 2013] fails because the involved homomorphic encryption system is invalid in the context of the scheme. This is due to that the general arithmetic over the field$\mathbb {R}$(the outsourced linear equations are constrained to the field) is not compatible with the modular arithmetic over any finite domain (the homomorphic encryption is constrained to the domain).

In PKC'2004, a constant-round group key exchange protocol is proposed. It takes advantage of the properties of the ElGamal encryption scheme combined with the standard secret sharing technique. In this paper, we show that the scheme can be greatly simplified because the standard secret sharing technique does not work well. Based on the observation, we present a revisited scheme. It saves about 4/5 cost of the original scheme. In addition, its security can be directly reduced to the underlying Schnorr's signature scheme and the ElGamal encryption.