[grayshirts]

If you are talking about ML then: Bayesian stuff, Variational and MCMC, Graphical models, Deep Learning, Linear Algebra, Probability theory, Statistics, Multivariate Calculus, Optimizations - as a baseline. I think the "average" person on HN would not be interested in ML if they had to learn it rigorously. It's much easier to blackbox the entire thing and label yourself as a "ML Engineer" then learn the fundamentals listed above. Having a Stats/Applied Math background would be very helpful.