Coining a new term in the title of your paper -- while defining it there or not -- is definitely something that you are permitted to do (so it is "proper", as you ask). I have seen this thing happen: the first example that sprung (!) to my mind was
Duchin, Moon Curvature, stretchiness, and dynamics. In the tradition of Ahlfors-Bers. IV, 19–30,
Contemp. Math., 432, Amer. Math. Soc., Providence, RI, 2007.
[First sentence of the MathReviews description of the paper:] The author introduces stretchiness, a new and interesting notion describing a kind of curvature of a metric space.
On the other hand, not only is this the first use of "stretchiness" in all of MathReviews, it is also the last.
A recent paper of mine is called "Quadratic reciprocity in abstract number rings", and you have to read the paper (or at least the abstract) to learn what an abstract number ring is.
Whether you should use a new term in the title of the paper is a very subjective and personal choice. I guess it's no secret that many, many people instinctively roll their eyes a little when they encounter a new word being coined. (In fact, I this to be an interesting socio-linguistic phenomenon. But moving on...) Making up your own language can make you look arrogant, overly playful, simply ignorant of existing nomenclature, and so forth. But the majority of academic terminology was coined by some prior academic within the relatively recent past rather than the mists of prehistory: as new ideas and new objects are created, we need new words for them.
Maybe -- maybe -- a good rule of thumb for this is that if you don't feel acutely that something would be lost upon removing your new term from the title, then it really doesn't need to be there. If you introduce the terminology within the paper itself, then you give the community a chance to chew on it, and then others can decide whether full promotion of the new term is a good idea. In fact, if I apply this standard to my own paper above, I am coming around to the idea that the title could be changed. The concept of an "abstract number ring" is not as important to me now as it was when I first started writing the paper.
Whether it is good or bad in your particular case seems impossible to answer without (i) seeing the title of your paper and (ii) having some real expertise and wisdom in the subject area of your paper. I would suggest that you find someone who satisfies (ii) -- e.g. a current or former advisor or mentor -- and show them your paper.