USACO Platinum. Because you can’t publish research in high school that anyone takes seriously unless you are genius material. And you need someone to guide you. It is too hard.
On the other hand if you do research that is good enough for an award at ISEF etc that is also good. You need guidance for this as well.
ISSN is short for International Standard Serial Number. A publication’s ISSN is a code that identifies that publication. Just like you wouldn’t assume a book is reputable/reliable just because it has an ISBN (International Standard Book Number), you should not assume a journal is reputable just because it has an ISSN.
Plenty of predatory journals and vanity presses have ISSNs, and indeed have most of the trappings of a reputable journal – but when they see a predatory journal on your list of ECs, knowledgeable college admissions officers will not be impressed. They won’t focus on the fact that you produced high-quality CS/stats research, even if you did. They’ll just think about the fact that you got scammed.
The USACO, on the other hand, is widely respected. So are ISEF and other similar programs, as @neela1 pointed out. I’d keep your focus there if I were you.
Both are equally good if the journal is really reputable and not just one with ISSN (don’t say ISSN number please, unless you also like saying ATM machine).
Keep in mind that publishing in a journal usually takes a lot of time with many back-and-forths. If I personally have to judge if a paper is legit, I’d spend time to actually read it. I doubt that AOs have that time. USACO Platinum has a known value.
On the other hand, published research can also be used for ISEF, STS, etc. So, yeah, it depends, sorry.