%0 Computer Program %A Dekker, Florine %A Erkin, Zekeriya %A Conti, Mauro %D 2025 %T Source code underlying the publication: Topology-Based Reconstruction Prevention for Decentralised Learning %U %R 10.4121/21572601.v2 %K reconstruction attacks %K statistical disclosure %K decentralised learning %K privacy-preserving summation %K graph girth %X <p>MATLAB code to reproduce results presented in the paper "Topology-Based Reconstruction Prevention for Decentralised Learning".</p><p><br></p><p>The source code is available as a git repository. Cached computation results are available as additional files.</p><p><br></p><p><strong>Git repository</strong></p><p>The git repository has two parts:</p><p><br></p><p>* <code>FeasibilitySim</code> contains the source code of the experiments from Section 4.4. Its main script is <code>FeasibilitySim.m</code>.</p><p>* <code>PerformanceSim</code> contains the source code of the experiments from Section 5.5. Its main script is <code>PerformanceSim.m</code>.</p><p><br></p><p>The parts are described in more detail in the file <code>README.md</code> in their respective folders.</p><p><br></p><p>Detailed usage instructions for both parts are available in the file <code>ARTIFACT-EVALUATION.md</code>.</p><p><br></p><p><strong>Files</strong></p><p>Cached computation results are optionally available, with filenames ending in <code>.cache.zip</code>. These allow inspecting the exact outputs without the need to re-run the computationally intensive experiments. Detailed usage instructions are available in the file <code>ARTIFACT-EVALUATION.md</code> inside the git repository.</p><p><br></p> %I 4TU.ResearchData