>>11273783Reading lots of papers, brushing up on mathematics related to my research, trying to build algorithms described in papers to make sure I have a solid understanding of how they work, testing new ideas in TensorFlow and comparing them to known algorithms. You can be involved in this field without dealing with mathematics beyond basic calculus, probability, and linear algebra. One can build an image classification system in a couple lines of code using higher level api without caring about the mathematics behind it. Machine learning is just doing statistics with a large dataset so you’re going to need to deal with mathematics everyday. But the kind of mathematics you will be doing varies greatly depending on the problems you’re tackling and whether you care more about fundamental research or engineering.
