Skip to content

Machine learning-based algorithm #7

@louis-hildebrand

Description

@louis-hildebrand

Once the manual analyzer is more or less working, we can consider developing a machine learning-based tool. This might be something like an encoder-decoder network.

Ideally, the ML tool would be able to understand a wide range of languages with minimal changes (unlike the manual tool, which needs a lexer for each language). We should also evaluate its robustness to different source code transformations, its efficiency, etc. The most ideal model would be able to understand an arbitrary language without training: the user just submits text (assembly, Chinese, whatever) and the model figures out how to lex it on the spot.

Metadata

Metadata

Assignees

No one assigned

    Labels

    epicHigh-level goal

    Type

    No type

    Projects

    Status

    Product Backlog

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions