© Celeste Martearena

I’m Daniel Hernandez Diaz, a Research Scientist on theoretical Machine Learning at the Department of Statistics of Columbia University (Postdoc position). My C.V. can be found here. I research on different aspects of the problem of reduction of the dimensionality of sequential data. On this website you can find links to published research and code. More generally, I am interested in most areas of machine intelligence, especially Deep Learning, Reinforcement Learning and whatever beast dwells in between.

In a previous life, I was a high-energy physicist. I conducted research on some of the puzzles of the so called “fundamental” physics. What is Dark Matter? Why are there more particles than antiparticles? Why three generations of fundamental fields when only one seems to be necessary (for most things at least 😊)? Although I do not practice physics professionally anymore, I may muse about it.

When I read a technical book, I do summaries that are usually of use to my understanding. I will be posting those here, once I properly edit them. It is possible you find them useful, but please note that since these summaries are personal, they typically assume whatever knowledge I possess at the time I wrote them. For a typical example, see Chapter 2 of the Reinforcement Learning book by R. Sutton and A. Barto.

Other interests of mine, like music and photography can also be found in some dark corner of this page.

I can be reached at:

dh[4*7][4*8] at columbia dot edu

where you should perform the little computations inside the brackets.