Machine Learning for Asset Managers (Elements in Quantitative Finance)

Highly recommended, its a pleasure to read. I really like the section on denoising/detoning covariance matrices. The material on mutual information of features is great. The bit on feature selection using MDA is golden. There are loads of little tricks in here, just sprinkled in, like position sizing using the normal cumulative distribution function of the Sharpe ratio.For practical examples of some ideas in action, pyportfolioopt now has an implementation (HROpt) of the minimum variance portfolio using de Prado's algorithm. It makes a material difference.

Too theoretical, no practical examples. Equations not clearly explained. Like an amalgamation of his academic papers. One or two chapters are useful. Could be written much better.

This is a good broad-brush introduction to tools beyond the standard statistical tools commonly used in the asset management industry. It is concise enough to be able to read in one sitting.It is nice that the book carries Python scripts, however, they are terse, badly formatted and without comments.Explanations are sketchy, often leave out critical details (e.g. HRP, covariance shrinkage).

Best book I ever read

easy to understand and very pratical