It really depends what range you want it invisible from. "From the outside" is quite a large range of values. I'm assuming "orbit around the star" (quite hard) rather than "orbit around a nearby star" (much easier) or "orbit around the planet itself (impossible)".
If you want a natural effect, probably the best you can do is to hide it the same way that islands were hidden in tales of yore - in a cloud.
A planet is defined as an object which has cleared its orbit, so technically, just by cluttering the orbit, you make the planet literally disappear - there is no longer a planet in that orbit, by definition.
But there's still one hell of a big rock. From a large enough distance, it'd be enough to fill the planet's orbit with enough high-albedo "chaff" that the planet would not be comparatively noticeable to instruments either when crossing the sun ("eclipse"), or when reflecting it ("full moon").
The closer you need the non-detection, the more dust you need.
A light mist will conceal an island from ten miles out, but you need a pea-souper to conceal it from ten yards!
Now, downside is, the dust will make the star look really weird. "A star that age, with so much dust, such thickly-populated rings? How strange."
The dust will also affect the planet, reducing incoming sunlight, and falling constantly into the atmosphere in some brilliant and beautiful eternal aurorae, perhaps laced with shooting stars of the larger particles. The planet would be better protected against radiation, and could possibly be heated by the aurorae at least to some extent, but I'm not sure how effective that would be.
There'd also need to be something to produce the dust, which might in turn be detectable.
[Edit: If you only need to protect from viewing directly in line with the planet's orbital plane, then you don't need dust in the planet's path - you can have a larger ring outside the planet, like our asteroid belt but much thicker, which would obscure it sufficiently, from that one angle.]
[Edit2: Note that ring thickness is unstable, but rings can be cabled, and if we're assuming that the rings can be produced, they can be renewed anyway. A solid ring (Dyson ring) resolves the stability problem but becomes a world more complex to create, and creates its own stability issues.
And all this is moot in the longer term anyway, since as @2012rcampion commented on another answer, with sufficiently sensitive instrumentation, and long enough study (at least one year, poss many more), merely observing the motion of any visible planets and the sun should be enough to find a planet in the mist even from another star.]