There are separate issues.

1. "Create" a lens - let's interpret that to be a design act - designing the curve according to focal length, required dispersion, etc. Yes, that's doable. There's lens design software around that will take a set of optimization criteria and some constraints and spit out a set of curves and glass types. Rustle up specific software recommendations on the amateur telescope making lists, I know I've seen them there.
2. Mold a lens - not without herculean effort. The homogeneity required of optical glass is, according to every thread on the topic I've been able to find, pretty much out of reach for garage tinkerer types. This isn't to permanently dissuade you from trying - I'd love few things more than to see another "you can't do this at home" techno-cultural touchstone falling. But you'll have to get (and maintain!) sufficiently transmissive glass, of sufficiently controlled optical parameters (index of refraction, dispersion), completely free of bubbles, and completely homogeneous. These are partially conflicting requirements, as it's difficult to, for example, have a thoroughly homogeneous melt without stirring, which can easily introduce bubbles.
3. Grind a lens - Definitely. I can't find the link now, but I saw an argument once that grinding (and, by extension, polishing) a good lens is easier than grinding a mirror, since the curves for a given f ratio are steeper with a lens, so errors show up more prominently during testing.
4. Polishing a lens - Yes again, as at this point it's the same as polishing a mirror surface, except perhaps for the steeper curve. Figuring an aspherical lens surface might be harder than figuring an aspherical mirror surface, for the same reason testing a lens is easier: the curves are steeper on a lens, so the lap has to flow a lot more as it mates to the lens in various positions.

Disclaimer: This is mostly book-knowledge, not so much hands-on knowledge. While I've ground and just started polishing a small mirror, I've never tried making a lens.