---
license: mit
pretty_name: Posture And Sustainment Optimization
task_categories:
- tabular-classification
- reinforcement-learning
language:
- en
tags:
- benchmark
- synthetic-data
- optimization
- logistics
- sustainment
- reinforcement-learning
- operations-research
size_categories:
- n<1K
configs:
- config_name: exp1_metrics
  default: true
  data_files:
  - split: train
    path: viewer/exp1_metrics.jsonl
- config_name: exp1_sensitivity
  data_files:
  - split: train
    path: viewer/exp1_sensitivity.jsonl
- config_name: exp1_significance
  data_files:
  - split: train
    path: viewer/exp1_significance.jsonl
- config_name: exp1_variance_decomposition
  data_files:
  - split: train
    path: viewer/exp1_variance_decomposition.jsonl
- config_name: sustainbench_categories
  data_files:
  - split: train
    path: viewer/sustainbench_categories.jsonl
- config_name: file_manifest
  data_files:
  - split: train
    path: viewer/file_manifest.jsonl
---

# Posture And Sustainment Optimization

This dataset contains research artifacts for **Posture and Sustainment Optimization**, a benchmark and simulation project for optimizing distributed posture, readiness, and sustainment decisions under uncertainty.

Source repository: https://github.com/anote-ai/research-postureandsustainmentoptimization

## Displayable Configs

The Hugging Face viewer reads normalized JSONL tables under `viewer/`:

- `exp1_metrics`: greedy placement baseline metrics across time steps.
- `exp1_sensitivity`: scenario-weighted readiness under uniform and skewed threat distributions.
- `exp1_significance`: statistical tests and effect sizes reported for Experiment 1.
- `exp1_variance_decomposition`: two-level variance decomposition for scenario-weighted readiness.
- `sustainbench_categories`: benchmark scenario category overview from the design document.
- `file_manifest`: source artifact paths, sizes, and SHA-256 hashes.

## Source Artifacts

The original project artifacts remain available in this dataset repo:

- `Research-PostureAndSustainmentOptimization-data.zip`: bundled source/data artifact from the repository.
- `source/`: README, design document, source code, scripts, tests, and package metadata.
- `paper/`: LaTeX paper source and PDF artifact.
- `results/`: original LaTeX result tables and Experiment 1 figure files.

## Use Notice

These artifacts are synthetic/offline research outputs for reproducibility and academic evaluation. They should not be interpreted as operational recommendations without independent validation.

## License

MIT. See the source repository for details.