CoInfoSim - Published Research Reports and Datasets

CoInfoSim is a research simulator for evaluating predictive cooperation across attribute subsets in supervised classification.

Install CoInfoSim

pip install coinfosim

Quick start

coinfosim scenario list
coinfosim scenario run occupancy --mode smoke

Available datasets

Files are mirrored byte-for-byte on this site; hashes are pinned in the installed package and verified before every run. See the machine-readable manifest.

Occupancy Detection

DOI: 10.24432/C5X01N

License: CC BY 4.0 (license text)

Creative Commons Attribution 4.0 International. Attribution required.

Source: https://archive.ics.uci.edu/dataset/357/occupancy+detection

Candanedo, L. (2016). Occupancy Detection [Dataset]. UCI Machine Learning Repository. https://doi.org/10.24432/C5X01N

  • datatraining.txt sha256:b2c4d0ce2b9e4e453c476f7125ef31aeec2d1f5c7f5572d0e80de3df6521ab56 (596,674 bytes)
  • datatest.txt sha256:1b92c7c1b2838963464fa891a610cf3c5db4becb7189189b29b330107a584c7f (200,766 bytes)
  • datatest2.txt sha256:d026d1bd5aeccd4aff4f3b3710d48e40613bd5fc370db7e61bbdcaa50d985095 (699,664 bytes)

Air Quality

DOI: 10.24432/C59K5F

License: CC BY 4.0 (license text)

UCI displays this dataset as CC BY 4.0. Data are intended for research use; please cite the associated publication (De Vito et al., Sensors and Actuators B: Chemical, 2008).

Source: https://archive.ics.uci.edu/dataset/360/air+quality

De Vito, S. (2016). Air Quality [Dataset]. UCI Machine Learning Repository. https://doi.org/10.24432/C59K5F

  • AirQualityUCI.csv sha256:13277ae5d8581e80b7be09d47c7d3d06fe9b8e957078f2cf6e859f955e62f996 (785,065 bytes)

SUPPORT2

DOI: 10.3886/ICPSR02957.v2

License: No explicit redistribution license identified; public research dataset; source acknowledgment required

No explicit redistribution license was identified for SUPPORT2. It is a public research dataset; follow the acknowledgment policy of the original HBiostat dataset site when using this file.

Source: https://hbiostat.org/data/

Knaus WA, et al. The SUPPORT Prognostic Model. Vanderbilt University Department of Biostatistics (HBiostat). https://doi.org/10.3886/ICPSR02957.v2

  • support2.csv sha256:79621945edf2a5c8dc36359684ff356d3c6025e773ba4fefac26f865f7894c78 (3,141,732 bytes)

Published scenario reports

9 published scenario reports.

Occupancy Detection Baseline (fast)

Dataset: Occupancy Detection

Does training on single-Gaussian synthetic data preserve the cooperative advantages observed when classifiers are evaluated on real Occupancy data? Equivalently: which training distribution best preserves the cooperative structure observed under real-data evaluation in the Occupancy Detection dataset?

scenarios/000001_occupancy_baseline_fast/occupancy_baseline_scenario_report_fast_000001.html

Machine-readable: semantic manifest · provenance (JSON-LD)

Open scenario report

Occupancy Detection Baseline (full)

Dataset: Occupancy Detection

Does training on single-Gaussian synthetic data preserve the cooperative advantages observed when classifiers are evaluated on real Occupancy data? Equivalently: which training distribution best preserves the cooperative structure observed under real-data evaluation in the Occupancy Detection dataset?

scenarios/000002_occupancy_baseline_full/occupancy_baseline_scenario_report_full_000002.html

Machine-readable: semantic manifest · provenance (JSON-LD)

Open scenario report

Occupancy Detection Baseline (smoke)

Dataset: Occupancy Detection

Does training on single-Gaussian synthetic data preserve the cooperative advantages observed when classifiers are evaluated on real Occupancy data? Equivalently: which training distribution best preserves the cooperative structure observed under real-data evaluation in the Occupancy Detection dataset?

scenarios/000000_occupancy_baseline_smoke/occupancy_baseline_scenario_report_smoke_000000.html

Machine-readable: semantic manifest · provenance (JSON-LD)

Open scenario report

SUPPORT2 180-Day Mortality Baseline (full)

Dataset: SUPPORT2

To what extent do class-conditional Single Gaussian and GMM synthetic training distributions preserve the cooperative channel-subset structure observed for 180-day mortality prediction on a fixed real SUPPORT2 test set?

scenarios/000007_support2_baseline_full/support2_scenario_report_full_000007.html

Machine-readable: semantic manifest · provenance (JSON-LD)

Open scenario report

SUPPORT2 180-Day Mortality Baseline (full)

Dataset: SUPPORT2

To what extent do class-conditional Single Gaussian and GMM synthetic training distributions preserve the cooperative channel-subset structure observed for 180-day mortality prediction on a fixed real SUPPORT2 test set?

scenarios/000008_support2_baseline_full/support2_scenario_report_full_000008.html

Machine-readable: semantic manifest · provenance (JSON-LD)

Open scenario report

UCI Air Quality Baseline (full)

Dataset: UCI Air Quality

To what extent do single-Gaussian and class-conditional GMM synthetic training distributions preserve the cooperative channel structure observed when classifiers detect elevated benzene concentration on a fixed future real-data evaluation period?

scenarios/000005_air_quality_baseline_full/air_quality_scenario_report_full_000005.html

Machine-readable: semantic manifest · provenance (JSON-LD)

Open scenario report

UCI Air Quality Baseline (full-scale)

Dataset: UCI Air Quality

To what extent do single-Gaussian and class-conditional GMM synthetic training distributions preserve the cooperative channel structure observed when classifiers detect elevated benzene concentration on a fixed future real-data evaluation period?

scenarios/000006_air_quality_baseline_full-scale/air_quality_scenario_report_full-scale_000006.html

Machine-readable: semantic manifest · provenance (JSON-LD)

Open scenario report

UCI Air Quality Baseline (smoke)

Dataset: UCI Air Quality

To what extent do single-Gaussian and class-conditional GMM synthetic training distributions preserve the cooperative channel structure observed when classifiers detect elevated benzene concentration on a fixed future real-data evaluation period?

scenarios/000003_air_quality_baseline_smoke/air_quality_scenario_report_smoke_000003.html

Machine-readable: semantic manifest · provenance (JSON-LD)

Open scenario report

UCI Air Quality Baseline (smoke)

Dataset: UCI Air Quality

To what extent do single-Gaussian and class-conditional GMM synthetic training distributions preserve the cooperative channel structure observed when classifiers detect elevated benzene concentration on a fixed future real-data evaluation period?

scenarios/000004_air_quality_baseline_smoke/air_quality_scenario_report_smoke_000004.html

Machine-readable: semantic manifest · provenance (JSON-LD)

Open scenario report

Machine-readable artifacts

Citation and license

See the project repository for the CoInfoSim citation file and license. Dataset citations and license/acknowledgment status are listed with each dataset above; SUPPORT2 in particular carries no open redistribution license and requires source acknowledgment rather than attribution under an open license.

Updated 2026-07-24 11:08:27Z