Sergey Yekhanin, Zinan Lin, Thomas Humphries
We lifted 9 functions out of this paper's own repositories and ran 9 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| t3humphries/PE-means | canonical | 9 of 9 |
| Function | Status | Where it lives |
|---|---|---|
| compute_epsilon | Ran | t3humphries/PE-means/src/DP_comp.py code served (permissive licence) · get_code("e2d6f607fcde45e2") |
| dataset_preprocess | Ran | t3humphries/PE-means/src/helpers.py code served (permissive licence) · get_code("d9155bf332092151") |
| delta_Gaussian | Ran | t3humphries/PE-means/src/DP_comp.py code served (permissive licence) · get_code("32fa80153875de17") |
| eps_Gaussian | Ran | t3humphries/PE-means/src/DP_comp.py code served (permissive licence) · get_code("773e62a14de7104f") |
| get_labels_memory_efficient | Ran | t3humphries/PE-means/src/pe_means_main.py code served (permissive licence) · get_code("fa200e45ebd8daf9") |
| levy_mutation | Ran | t3humphries/PE-means/src/pe_means_main.py code served (permissive licence) · get_code("27d2fedf6b91f95c") |
| load_txt | Ran | t3humphries/PE-means/src/helpers.py code served (permissive licence) · get_code("2a9d75d1feca994c") |
| run_single_iteration | Ran | t3humphries/PE-means/baselines/icml_baseline.py code served (permissive licence) · get_code("76a4d233387691ad") |
| simple_loss | Ran | t3humphries/PE-means/src/helpers.py code served (permissive licence) · get_code("9ffd334fedc7e7c3") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
We study the problem of differentially private (DP) 𝑘-means clustering in Euclidean space. Previous solutions rely on summing the private data directly, which induces a sensitivity proportional to the domain. We introduce PE-means, an extension of the private evolution (PE) algorithm (an increasingly popular method for synthetic data generation), to the problem of 𝑘-means clustering. The key advantage of PE is that it only computes a private histogram with constant sensitivity to guide the evolution. Our adaptation of PE includes new evolutionary operators for clustering, as well as other algorithmic improvements of independent interest. Overall, PE-means achieves an average improvement of 26% in clustering loss over state-of-the-art baselines such as Google's LSH-based algorithm and DP-Lloyd variants.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2606.00342")
get_code_for_paper("2606.00342")
have("2606.00342")
Connect an agent — have() is free.