There are 3 core elements to invention: ideation, prototyping, and validation. Invention typically is an iterative cycle, where the inventor is trying to solve a problem for a customer, and create a solution that is 10x better than the competition.
Ideation involves gathering and evaluating data, both from life experience, and then from the protoyping and validation steps (once you have gone through the invention cycle once). This is arguably the most fun part, but if all you do is stick in the ideation phase, all you will be able to do is say "man, I had that idea years before the Winklevoss brothers." One way to improve your ideation is to focus on ideas that you can prototype and validate, either yourself or with the help of an ally or expert.
Prototyping is about transitioning your ideas into a usable form. This can be creating a physical prototype for objects, a software prototype for apps and services, or even lower-fidelity prototypes (screenshots/mockups/videos) that will allow you to learn more about how the invention works OUTSIDE of your head.
Validation is the most important step with invention, and is also the most frequently skipped step by inventors. This step is about getting feedback from the actual customer who is going to buy your product or service. Getting good validation is an art form, because you have to distinguish strategic feedback (feedback that hits the core of whether your invention is heading in the right direction towards profitably solving a customer's pain) v tactical feedback (colors, finish, details...the polish that will make a good product great, but is generally not helpful during the invention process).