experiment_01_acs.md

Experiment #1: ACS patients flow with classification

Implemented in experiment_01_acs.py

Experiment description

Simulation of patients with two flows: ACS patients from ACS clustered data and background flow for surgery facilities load with surgery flow data. Parameters variation:

• Target (ACS) flow scaling: 0.5, 1.0, 1.5, 2.0
• Background (surgery) flow scaling: 0.5, 1.0, 1.5, 2.0
• Number of parallel surgeries facilities: 1, 2, 3, 4, 5

Every combination of three parameters is executed 100 times with simulation of 60 days (445*100=8.000 runs in total). Executed in parallel.

Simulation results

(a) - QQ-plot for total LoS for all cases and top-3 clusters; (b) - percentage of cases where waiting in queue was detected for different background scales; (c) mean waiting time for different background scales

Additional visualization (Jupyter notebooks):

• LoS plotting
• Queueing analysis and plotting
• Queueing analysis and plotting (old version)