Re:GitHub releases an ImageNet for code and a CodeSearchNet ...
by YeGoblynQueenne on Friday September 27, 2019 @06:43AM
(#21090414)
>> Write some unit tests, a human description of what it does and based on the source code and description of existing software the system would basically “debug the program into existence” for you.
This is already possible, but not with deep learning which is probably the reason you haven't heard of it.
Learning programs from specifications (not necessarily in natural language) is the subject of the field of Program Synthesis [1].
Learning programs from examples of their input and outputs (which is basicaly writing unit tests) is the subject of Inductive Programming [2].
Inductive Programming encompasses the fields of Inductive Logic Programming and Inductive Functional Programming, that target logic and functional programming languages, respectively.
You won't find anything learning from examples in javascript or python directly, though- imperative languages are too sloppy for that sort of thing.
This is already possible, but not with deep learning which is probably the reason you haven't heard of it.
Learning programs from specifications (not necessarily in natural language) is the subject of the field of Program Synthesis [1].
Learning programs from examples of their input and outputs (which is basicaly writing unit tests) is the subject of Inductive Programming [2].
Inductive Programming encompasses the fields of Inductive Logic Programming and Inductive Functional Programming, that target logic and functional programming languages, respectively.
You won't find anything learning from examples in javascript or python directly, though- imperative languages are too sloppy for that sort of thing.
___________
[1] https://en.wikipedia.org/wiki/Program_synthesis
[2] https://en.wikipedia.org/wiki/Inductive_programming