4 ms·
It doesn't stop at forward-mode, and reverse-mode, because finding the optimal mode relates to a complicated graph problem, which, IIRC, is NP-hard in general.
by conistonwater 9y ago
It doesn't stop at forward-mode, and reverse-mode, because finding the optimal mode relates to a complicated graph problem, which, IIRC, is NP-hard in general. Not only that, but most of the difficulty is in getting these tools to work on arbitrary codebases, making it more a software engineering problem rather than a maths problem.
- vog 9y ago> It doesn't stop at forward-mode, and reverse-mode, because finding the optimal mode relates to a complicated graph problem, which, IIRC, is NP-hard in general. Yes, there are combined forward-reverse-mode approaches as well as "cross country" approaches. That's in part of what I menat with a "whole world to discover". > Not only that, but most of the difficulty is in getting these tools to work on arbitrary codebases, making it more a software engineering problem rather than a maths problem Depending on your viewpoint, it is neither SE nor math, but a PL (programming languages) issue, meaning that viewing the program e.g. as register machine (classic assembly) yields to different approaches than viewing it as lambda calculus expression (which provided great results, but it is hard to transport the results back to languages like Fortran or C++).