I have $N$ points ${P_1, P_2, P_3, ..., P_N}$ with dimension $K$, where $K\gg N$. Given another point $Q$ with the same dimension, I want to get which of is the point $P_n$ which is closest to $Q$. The solution is easy, but since I want to perform the operation multiple times and using the fact that $K\gg N$ I am looking for a way to transform ${P_1, P_2, P_3, ..., P_N}$ and $Q$ to a lower dimension which could give the same result with less storage and computation. Is this possible?
Thanks!!!!