A Maximum Entropy Method for Inverting Laplace Transforms of Probability Density Functions
提出一种最大熵方法，用于反演正随机变量密度函数的拉普拉斯变换，仅需少量变换值即可获得准确近似，并通过伽马、对数正态、逆高斯和帕累托分布验证。
This papers presents a maximum entropy method for inverting Laplace transforms of density functions of positive random variables. The maximum entropy density is very flexible and can assume a variety of different shapes. Accurate approximations to the true density can be obtained even when only a few transform values are available. Numerical evidence is provided for gamma, lognormal, inverse Gaussian and Pareto distributions.