The answer here depends so much on how much money and how much time and how much interest you have that I can't give you a simple formula. Nonetheless here are a few basic observations that might prove useful at varying levels of interest:
1. At some you should just start buying some stuff. You're going to make some mistakes at first, treat that as part of your learning curve and as part of the price of the broader endeavor.
2. Don't ever think you can make money buying and selling art. The bid-ask spread is a bitch, and finding the right buyers is a complex and time-consuming matching problem.
3. Art is strongly tiered in a hierarchical fashion. That means most fields are incredible bargains, at least relative to the trendy fields. A lot of HNW buyers are looking for large, striking contemporary works they can hang over their sofas in their second homes in Miami Beach or Los Angeles or Aspen. Good for them, as many of those works are splendid. Nonetheless that opens up opportunities for you. I find the price/beauty gradient ratio can be especially favorable for textiles, ethnographic works, Old Master drawings (and sometimes paintings), paintings from smaller or obscure countries, various collectibles, and many other areas.
4. As for the price/beauty gradient, prints, lithographs, and watercolors usually are much cheaper than original paintings. And they are not necessarily of lower quality. Figure out early on which are the artists whose prints can be as good or better than many of their original works (Lozowick, Picasso, and Johns would be a few nominees) and learn lots about those areas. A Goncharova painting can costs hundreds of thousands or millions of dollars, but one of her Ballet Russe designs -- an original done by the same hand -- can go for thousands.
4b. The "mainstream art market" still discriminates in favor of "original" works, but it already has started laying this convention aside for photographs, and I wonder if further erosion along these lines is not on the way. The "internet generation" is getting wealthier all the time, and do they all hate reproducibility so much?
5. Pick a small number of areas and specialize in them. Learn everything you can about them. Everything. Follow auction results. Read about their history. Read biographies of their creators. Go visit exhibitions. And so on. It is also a great way to learn about the world more generally.
6. If you are an outsider, you can't just walk up to a gallery and buy the best stuff at a market clearing price. You have to invest in your relationships there. Or consider buying at auction. Whatever your choice, be aware of the logic and why things work that way. Selling practices are also an exercise in reputation management of the artist and of the gallery, and maybe they think you are not up to snuff as a buyer!
7. Visit other people's art collections as much as you can. You will learn a great deal this way, and learn to spot new forms of foolishness that you had never before imagined.
8. Don't treat art collecting as like shopping, or as motivated by the same impulses. If you do, you will accumulate a lot of junk very quickly. Thinking of it more as building out a long term narrative of what an artistic field, and a culture, is all about. Fine if you don't want to do that! It is a demanding exercise, and if you wish to escalate your collecting to higher levels, you need to ask yourself if you are really up for that. Does it sound like something you would be good at?
9. Fakes are rampant in so many parts of the art world, but they are especially likely if the artist is "popular" (e.g., Chagall, Dali) or if the style is easily copied ex post (Malevich). In contrast, if you buy a piece of complex stained glass, it is probably the real thing. The major auction houses are usually reasonably good at rooting out fakes, but there is no institution you can trust 100 percent. And sometimes, as with the recently auction Botticelli and da Vinci paintings, no one really knows for sure (Botticelli pro and con; in any case I don't like the work, certainly not for $45 million).
10. Don't buy art on the basis of the artist's name. This is a good way to end up with a lot of crap and, for that matter, fakes. Just about every famous artist has a fair number of mediocre works, overpriced at that. (That said, if you really just want to "collect names," you will find it is remarkable on a limited income just how many top names you can wrack up.)
11. Few of the important art collections were built by just throwing tons of money at the task. That is a recipe for being ripped off, and it attracts poor quality sellers to your orbit. You have to understand something more deeply than other people do. Obviously money helps, but you can't rely on outbidding others as your most important ally.
12. Maybe sometime I'll tell you the story of how I obtained an especially fine, rare work by throwing a stone at a wild dog in rural Mexico. Or how I tracked another painter down at the mental hospital.
There is much more I could say of relevance (e.g., how to present yourself to dealers? how to avoid winner's curse?), but I'll stop there for now.