Discrete Approximations of Continuous Distributions by Maximum Entropy

Published in Economics Letters, 2013

When I was working on my job market paper in 2012, I had to solve a numerical example and needed to compute some expectations that do not have closed-form expressions. Believe it or not, in those days I didn’t know about numerical integration, and instead of googling, I came up with a method to assign probabilities on grid points to match moments by applying the maximum entropy principle, which I was familiar. I approached my former classmate Ken’ichiro Tanaka, who is an expert on numerical analysis, and we wrote a paper together.

Direct Link