Estimating causal effects with optimization-based methods: A review and empirical comparison

In the absence of randomized controlled and natural experiments, it is necessary to balance distributions (observable) covariates treated control groups in order obtain an unbiased estimate a causal effect interest; otherwise, different size may be estimated, incorrect recommendations given. To achieve this balance, there exist wide variety methods. particular, several methods based on optimization models have been recently proposed inference literature. While these optimization-based empirically showed improvement over limited number other their relative ability effects, they not thoroughly compared each noteworthy addition, we believe that unaddressed opportunities operational researchers could contribute with advanced knowledge optimization, for benefits applied use tools. review paper, present overview literature describe more detail methods, provide comparative analysis prevailing discuss new