自9月份开始初次做市，至今已过去完整的一个月，有一些体会和反思，记录在这里。（以下内容并不涉及具体策略）

This post summarizes several methods in calculating distance-to-default (DD), which is an important factor in measuring a firm’s probability of default.

In this post, I will try to explain what the idea of gradient descent is, which is the basis of a lot of machine learning algorithms, especially neural networks. I will try to explain this in the easiest way, but feel free to grab your long-lost calculus book along the way. :)

In short, my answer is around two months from now. I will explain in details.

Copyright of this article belongs to SoYoga (文章版权归如是瑜伽所有). If images cannot be loaded, please visit this link instead.

This is part of a Moody’s document, Modeling Credit Portfolios, explanining RiskFrontier methodology.