3 |
3 |
4 This package contains [Rust][] codes for the manuscript “_Proximal methods for |
4 This package contains [Rust][] codes for the manuscript “_Proximal methods for |
5 point source localisation_” by Tuomo Valkonen ⟨tuomov@iki.fi⟩. It concerns |
5 point source localisation_” by Tuomo Valkonen ⟨tuomov@iki.fi⟩. It concerns |
6 solution of problems of the type |
6 solution of problems of the type |
7 $$ |
7 $$ |
8 \min_{μ ∈ ℳ(Ω)}~ F(x) + λ \|μ\|_{ℳ(Ω)} + δ_{≥ 0}(x), |
8 \min_{μ ∈ ℳ(Ω)}~ F(μ) + λ \|μ\|_{ℳ(Ω)} + δ_{≥ 0}(μ), |
9 $$ |
9 $$ |
10 where $F$ is a data term, and $ℳ(Ω)$ is the space of Radon measures on the |
10 where $F$ is a data term, and $ℳ(Ω)$ is the space of Radon measures on the |
11 (rectangular) domain $Ω ⊂ ℝ^n$. Implemented are $F(x)=\frac12\|Ax-b\|_2^2$ and |
11 (rectangular) domain $Ω ⊂ ℝ^n$. Implemented are $F(μ)=\frac12\|Aμ-b\|_2^2$ and |
12 $F(x)=\|Ax-b\|_1$ for the forward operator $A \in 𝕃(ℳ(Ω); ℝ^m)$ modelling a |
12 $F(μ)=\|Aμ-b\|_1$ for the forward operator $A \in 𝕃(ℳ(Ω); ℝ^m)$ modelling a |
13 simple sensor grid. For the 2-norm-squared data term implemented are the |
13 simple sensor grid. For the 2-norm-squared data term implemented are the |
14 algorithms μFB, μFISTA, and μPDPS from the aforementioned manuscript along with |
14 algorithms μFB, μFISTA, and μPDPS from the aforementioned manuscript along with |
15 comparison relaxed and fully corrective conditional gradient methods from the |
15 comparison relaxed and fully corrective conditional gradient methods from the |
16 literature. For the 1-norm data term only the μPDPS is applicable. |
16 literature. For the 1-norm data term only the μPDPS is applicable. |
17 |
17 |