Graphical Modelling and Causal Inference
Published 1989 | Version v3
Dataset Open

ALARM (A Logical Alarm Reduction Mechanism)

Creators

Description

This synthetic dataset represents alarm message system for patient monitoring. It was introduced in Beinlich et al. (1989).

 

Task: The dataset collection can be used to study causal discovery algorithms.

 

Summary: 

  • Size of collection: 20,000 x 37
  • Task: Causal Discovery Problem
  • Data Type: Categorical Data
  • Dataset Scope: Standalone Dataset
  • Ground Truth: Known Graph
  • Temporal Structure: Static Data
  • License: CC0 (generated for bnlearn)
  • Missing Values: No Missing Values

 

Missingness Statement: There are no missing values.

 

Collection: 

The alarm dataset contains the following 37 variables:

  • CVP: Central venous pressure (LOW / NORMAL / HIGH)
  • PCWP: Pulmonary capillary wedge pressure (LOW / NORMAL / HIGH)
  • HIST: History (TRUE / FALSE)
  • TPR: Total peripheral resistance (LOW / NORMAL / HIGH)
  • BP: Blood pressure (LOW / NORMAL / HIGH)
  • CO: Cardiac output (LOW / NORMAL / HIGH)
  • HRBP: Heart rate / blood pressure (LOW / NORMAL / HIGH)
  • HREK: Heart rate measured by an EKG monitor (LOW / NORMAL / HIGH)
  • HRSA: Heart rate / oxygen saturation (LOW / NORMAL / HIGH)
  • PAP: Pulmonary artery pressure (LOW / NORMAL / HIGH)
  • SAO2: Arterial oxygen saturation (LOW / NORMAL / HIGH)
  • FIO2: Fraction of inspired oxygen (LOW / NORMAL)
  • PRSS: Breathing pressure (ZERO / LOW / NORMAL / HIGH)
  • ECO2: Expelled CO2 (ZERO / LOW / NORMAL / HIGH)
  • MINV: Minimum volume (ZERO / LOW / NORMAL / HIGH)
  • MVS: Minimum volume set (LOW / NORMAL / HIGH)
  • HYP: Hypovolemia (TRUE / FALSE)
  • LVF: Left ventricular failure (TRUE / FALSE)
  • APL: Anaphylaxis (TRUE / FALSE)
  • ANES: Insufficient anesthesia/analgesia (TRUE / FALSE)
  • PMB: Pulmonary embolus (TRUE / FALSE)
  • INT: Intubation (NORMAL / ESOPHAGEAL / ONESIDED)
  • KINK: Kinked tube (TRUE / FALSE)
  • DISC: Disconnection (TRUE / FALSE)
  • LVV: Left ventricular end-diastolic volume (LOW / NORMAL / HIGH)
  • STKV: Stroke volume (LOW / NORMAL / HIGH)
  • CCHL: Catecholamine (NORMAL / HIGH)
  • ERLO: Error low output (TRUE / FALSE)
  • HR: Heart rate (LOW / NORMAL / HIGH)
  • ERCA: Electrocauter (TRUE / FALSE)
  • SHNT: Shunt (NORMAL / HIGH)
  • PVS: Pulmonary venous oxygen saturation (LOW / NORMAL / HIGH)
  • ACO2: Arterial CO2 (LOW / NORMAL / HIGH)
  • VALV: Pulmonary alveoli ventilation (ZERO / LOW / NORMAL / HIGH)
  • VLNG: Lung ventilation (ZERO / LOW / NORMAL / HIGH)
  • VTUB: Ventilation tube (ZERO / LOW / NORMAL / HIGH)
  • VMCH: Ventilation machine (ZERO / LOW / NORMAL / HIGH)

 

Files: 

  • alarm.csv: dataset
  • alarm.bif: Bayesian Network from (Scutari (2010), License CC BY-SA 3.0). The network was used for data generation in Beinlich et al. (1989).

Files

alarm.csv

Files (5.7 MB)

Name Size Download all
md5:36d1b19ed6a4514e17d2321d1ffb580a
13.4 kB Download
md5:0df128ae98e2a899c9de1f27c7c49d4a
5.7 MB Preview Download

Additional details

Identifiers

DOI
10.1007/978-1-4614-6446-4

Related works

Is described by
Computational notebook: 10.5281/zenodo.15304389 (DOI)
Is documented by
Book chapter: 10.1007/978-1-4757-3502-4_6 (DOI)

References

  • Beinlich I, Suermondt HJ, Chavez RM, Cooper GF (1989). "The ALARM Monitoring System: A Case Study with Two Probabilistic Inference Techniques for Belief Networks". Proceedings of the 2nd European Conference on Artificial Intelligence in Medicine, 247–256.
  • Scutari M (2010). "Learning Bayesian Networks with the bnlearn R Package." Journal of Statistical Software, 35(3), 1–22. doi:10.18637/jss.v035.i03