I’m pursuing ML research outside my job (ML related too) and I have results fit for submission to relevant journals/conferences. The problem is that these require some kind of official affiliation, usually linked to an academic/company email address.<p>Given that all the work has been done outside my work hours (and more importantly on hardware of my own), I don’t see a reason to publish these with my employer affiliation. I could use my old affiliation from graduate school times but again this doesn’t seem right.<p>I am playing with the idea of buying a .org domain and setup a “research institute”. Is that simple? For reference, I’m in UK.
I saw a paper signed by an "Independent Researcher" that was probably a polite way to say "Currently Unenployed". Perhaps you can try that.<p>Writing papers it's difficult, becuase there are a lot ot t to cross and i to dot. Also, the result must not only be useful, but also must be "interesting" for whatever definition of "interesting" the editor of the journal have.<p>It's much easier if you can colaborate with a group that is alreading publishing.
<i>I could use my old affiliation from graduate school times but again this doesn’t seem right.</i><p>Does it feel less right than using an only-on-paper "research institute?"<p>[Scare quotes in original.]<p>Good luck.