Estimating individual treatment effect: generalization bounds and algorithms

Benchmark Model Rank Results
causal-inference-on-ihdpCounterfactual Regression + WASS#5Average Treatment Effect Error: 0.27
causal-inference-on-ihdpTARNet#6Average Treatment Effect Error: 0.28
causal-inference-on-ihdpCausal Forest#7Average Treatment Effect Error: 0.4
causal-inference-on-ihdpBalancing Neural Network#8Average Treatment Effect Error: 0.42
causal-inference-on-ihdpk-NN#10Average Treatment Effect Error: 0.79
causal-inference-on-ihdpBalancing Linear Regression#11Average Treatment Effect Error: 0.93
causal-inference-on-ihdpRandom Forest#12Average Treatment Effect Error: 0.96
causal-inference-on-jobsCFR MMD#4Average Treatment Effect on the Treated Error: 0.08
causal-inference-on-jobsCFR WASS#5Average Treatment Effect on the Treated Error: 0.09