A Held-Out Site Is a Sample of One: What Random, Spatial and Site-Disjoint Splits Each Estimate, Why Leakage That Differs Between Configurations Reorders a Comparison, and How Many Sites a General Claim Needs

APA

Mahendrakar, P. (2026). A Held-Out Site Is a Sample of One: What Random, Spatial and Site-Disjoint Splits Each Estimate, Why Leakage That Differs Between Configurations Reorders a Comparison, and How Many Sites a General Claim Needs [Preprint]. Life of Research. https://doi.org/10.5281/zenodo.23824501

Copy APA

BibTeX

@misc{lor1702026,
  title        = {A Held-Out Site Is a Sample of One: What Random, Spatial and Site-Disjoint Splits Each Estimate, Why Leakage That Differs Between Configurations Reorders a Comparison, and How Many Sites a General Claim Needs},
  author       = {Pranay Mahendrakar},
  year         = {2026},
  publisher    = {Life of Research},
  doi          = {10.5281/zenodo.23824501},
  url          = {https://lifeofresearch.com/preprints/170},
  keywords     = {spatial cross-validation, data leakage, site-disjoint evaluation, model comparison, domain generalization, agricultural robotics, evaluation methodology, clustered inference},
  note         = {LOR-00170}
}

Copy BibTeX More on the record page