Existing methods for solving inverse problems, such as the regularization method, look mostly for a quasi-solution that may not be a solution to the original problem, but only approximate it. The technique proposed in this article allows us to describe the set of solutions that satisfy the problem and find the solution that best fits the existing a priori information about the distribution of model properties based on these equivalent solutions.

中文翻译：

---

求解线性方程组的投影方法：重力测量应用

现有的求解逆问题的方法，例如正则化方法，主要寻找可能不是原始问题的解而只是近似它的拟解。本文中提出的技术使我们能够描述满足问题的解决方案集，并基于这些等效解决方案找到最适合现有模型属性分布的先验信息的解决方案。