Rapid Analytics & Model Prototyping
using Python
RAMP data challenges are a great way to collaboratively prototype and benchmark machine learning workflows. Participants submit their predictive solution (code), competing for the best score, and provide organisers with fully-functioning prototypes.
Solutions submitted by participants are made publicly available during the 'open phase', at the end of a challenge. Participants can reuse and learn from each other's solutions, accelerating development.
RAMP workflow allows 'model blending'. A blend of the best models often achieves a better score than the best single submission.
RAMP data challenges are a great tool to learn data science! Participants can learn from each other, especially during the 'open phase'.
RAMP packages offer versatile tools to define, build and optimize machine learning workflows via data challenges, great for prototyping models in your data science team.
RAMP workflow allows you to formalise a machine learning workflow by defining score types (metrics), workflow elements and prediction types.
RAMP board consists of a bundle of modules that allow you to easily setup your own RAMP server and deploy new data challenges.
RAMP was originally developed as a tool for data scientists to efficiently and collaboratively solve the data analytics aspect of high-impact domain science problems. Since then, RAMP has undergone many iterations and has been used in teaching and benchmarking machine learning algorithms as well.