Is the root of a sum of squared differences convex?

745 Views Asked by At

Let $x \in \mathbb{R}^n$. Let there be a collection of functions $d_i = (x_j - x_k)^2$ (note that the subscripts $j$ and $k$ are fixed for each $d_i$, and there can be repeated use of subscripts on different $d_i$'s. Define the function:

$$f(x) = \sqrt{ \sum_i d_i }$$

Is this function convex? I know in the case where subscripts are not repeated, it is:

Is a distance function on $\mathbb{R}^n$ convex?

I think this proof still holds, but I'd like it if someone could help me validate it.

The context is that I wish to formulate a convex optimization problem. I know I can always do this by adding auxiliary variables such that all the subscripts are unique, and then adding equality constraints of the form $x_a = x_b$ for any two auxiliary variables which are supposed to represent the same variable in the original problem. But, I'd like to know if the unconstrained formulation is convex.

1

There are 1 best solutions below

0
On BEST ANSWER

Let the vector of differences be given by $C x$, where $C \in \mathbb{R}^{m \times n}$ is the oriented incidence matrix of a graph with $n$ nodes and $m$ edges. Hence,

$$f (x) := \sqrt{x^T C^T C x} = \sqrt{\|Cx\|_2^2} = \| C x\|_2$$

where $C^T C$ is the Laplacian matrix of the same graph. Thus, $f$ is convex. It is strictly convex if $C$ has full column rank.