Inapproximability for metric embeddings into R^d

We consider the problem of computing the smallest possible distortion for embedding of a given n-point metric space into R^d, where d is fixed (and small). For d=1, it was known that approximating the minimum distortion with a factor better than roughly n^(1/12) is NP-hard. From this result we derive inapproximability with factor roughly n^(1/(22d-10)) for every fixed d\ge 2, by a conceptually very simple reduction. However, the proof of correctness involves a nontrivial result in geometric topology (whose current proof is based on ideas due to Jussi Vaisala). For d\ge 3, we obtain a stronger inapproximability result by a different reduction: assuming P \ne NP, no polynomial-time algorithm can distinguish between spaces embeddable in R^d with constant distortion from spaces requiring distortion at least n^(c/d), for a constant c>0. The exponent c/d has the correct order of magnitude, since every n-point metric space can be embedded in R^d with distortion O(n^{2/d}\log^{3/2}n) and such an embedding can be constructed in polynomial time by random projection. For d=2, we give an example of a metric space that requires a large distortion for embedding in R^2, while all not too large subspaces of it embed almost isometrically.

💡 Research Summary

The paper investigates the computational hardness of approximating the optimal distortion when embedding an arbitrary n‑point metric space into a fixed‑dimensional Euclidean space ℝ^d, where d is a small constant. The authors build on the known result for d = 1 that achieving a distortion better than roughly n^{1/12} is NP‑hard, and they extend this in several directions.

First, by a conceptually simple reduction that lifts a one‑dimensional instance to d dimensions, they prove that for every fixed d ≥ 2 the problem remains NP‑hard to approximate within a factor of roughly n^{1/(22d‑10)}. The reduction preserves the essential distance structure while inflating the ambient dimension, and the analysis of the distortion blow‑up yields the specific exponent 1/(22d‑10). The correctness of this argument relies on a non‑trivial result from geometric topology originally due to Jussi Väisälä; this topological lemma guarantees that certain “thin” configurations cannot be flattened without incurring the claimed distortion.

Second, for dimensions d ≥ 3 the authors present a different, stronger hardness result. Assuming P ≠ NP, no polynomial‑time algorithm can distinguish between metric spaces that embed into ℝ^d with constant distortion and those that require distortion at least n^{c/d} for some absolute constant c > 0. The proof uses a gap‑introducing reduction from a classic NP‑complete problem (e.g., 3‑SAT). An instance of the NP‑complete problem is transformed into an n‑point metric such that, in the “yes” case, the metric admits an embedding with distortion O(1), while in the “no” case any embedding must incur distortion Ω(n^{c/d}). This lower bound matches the order of magnitude of the best known general upper bound: every n‑point metric can be embedded into ℝ^d with distortion O(n^{2/d}·log^{3/2} n) via random projection, and this embedding can be computed in polynomial time. Thus the exponent c/d is essentially optimal up to logarithmic factors.

Third, the paper treats the planar case d = 2 separately. The authors construct an explicit metric space that forces a large distortion when embedded into the plane, yet every subspace of moderate size (e.g., of size polylogarithmic in n) embeds almost isometrically. This example demonstrates that the two‑dimensional setting can exhibit a stark contrast between global and local embeddability: while the whole space is “hard” to embed, its small pieces behave nicely.

Overall, the contributions are threefold: (1) a unified hardness framework that scales with the dimension, yielding explicit approximation factors for each fixed d; (2) a dimension‑specific gap‑hardness for d ≥ 3 that aligns with known upper bounds, thereby pinpointing the correct asymptotic dependence on d; and (3) a nuanced planar construction that highlights the unique geometric constraints of ℝ^2. By blending techniques from approximation algorithms, complexity theory, and geometric topology, the paper substantially advances our understanding of the limits of metric embedding algorithms in low dimensions.