The phrase "a somewhat different set" is correct and usable in written English.
It can be used when referring to a collection or group that has variations or distinctions compared to another set.
Example: "In this study, we analyzed a somewhat different set of data to compare the results with our previous findings."
Alternatives: "a slightly different group" or "a somewhat altered collection".
When all is said and done, Mr. DeLeo said, Delcon intends to emerge from bankruptcy ready to tackle a somewhat different set of projects.
Each town has a somewhat different set of construction rules, but they all aim to encourage the construction of "island friendly" houses.
More rationally, we develop our own interests, our own talents, or own set of friends, and stand behind a somewhat different set of standards.
Hallucinations in the waking state also would involve the same phenomenon, produced by a somewhat different set of psychological or physiological circumstances.
But it soon became clear that mice, though very similar to people at the genetic level, have a somewhat different set of ZIP codes, often using the same code but in different organs.
The point is that what Silver and Cohn and Wang — there's a law firm for you! — are doing is not just arriving at a somewhat different result by a somewhat different set of formulae, but pursuing an entirely different (no other word to use here) epistemology.
(See also [3, 4].) We will consider a somewhat different setting, as follows.
"There's also somewhat different set of securities that can be submitted," Wyss noted.
Moreover, the analysis of Segments 3 and 4 (using a somewhat different strain set) validated the data from initial data Segments 1 and 2. These measures reduce the likelihood that we inadvertently introduced a bias for certain kinds of genes that are more or less likely to have undergone recombination [ 27].
Two meta-analyses of somewhat different sets of these studies demonstrated that the difference in the birth sex ratios of high- and low-ranking females declined as the sample size increased [4], [5], suggesting that the variation represented stochastic effects of small samples.
While other groups have used overlapping, but somewhat different, sets of key residues (Alleyne et al., 2009), we find that including additional features beyond the set selected using MIp actually decreased performance for the SVM and KNN models, and only increased performance slightly for RF (see below).
