demo_syntheticDataExp.md

Usage of runSyntheticExp

Import all functions in runSyntheticExp

from runSyntheticExp import *

Initialization

Step 1: Specify the input population with size of user and protected group

user_N = 100
pro_N = 50

Step 2: Choose the fairness measure will be used in the experiment

• Choose from ["rKL", "rND", "rRD"].
• rKL represents KL-divergence fairness measure.
• rND represents normalized difference fairness measure.
• rRD represents ratio difference fairness measure.

gf_measure = "rKL"

Step 3: Set the cut point at where to compute the fairness measure

cut_point = 10

Step 4: Specify the file to output optimization results

output_fn = "Fairness_synthetic_"+gf_measure

Run fairness measure expetiments of synthetic data

main(user_N,pro_N,gf_measure,cut_point,output_fn)

print "Finished experiments on synthetic data"
print "Result stores in "+ output_fn+"_user"+str(user_N)+"_pro"+str(pro_N)+".csv"

Finished mixing proportion  0.0
Finished mixing proportion  0.1
Finished mixing proportion  0.2
Finished mixing proportion  0.3
Finished mixing proportion  0.4
Finished mixing proportion  0.5
Finished mixing proportion  0.6
Finished mixing proportion  0.7
Finished mixing proportion  0.8
Finished mixing proportion  0.9
Finished mixing proportion  0.98
Finished experiments on synthetic data
Result stores in Fairness_synthetic_rKL_user100_pro50.csv