Edge-ML is a library of algorithms based on the mathematical approch MODL . It constitues a pipe of Machine Learning, entirely automatised (auto ML) allowing training a classifier from numerical, categorical or sequential data (ex : web session, texts, logs…).
Edge-ML is a cutting edge technology which allows one to efficiently carry out the majority of the Machine Learning projects. Edge-ML offers a smarter practice of Machine Learning, more agile, more robust and while reducing the costs… The Data Scientist is a rare resource, Edge-ML increases its efficiency.
Natural trade-off between Performance and Reliability: The models have a precision close to the optimal and a warranted robustness, which securised the use of models.
Use of common hardware resources: Optimised exploitation of your own computers and possibility to learn the models directy on the devices (IOT, mobiles...).
Drastic reduction of the duration of the projects: Avoiding the cleaning of data and the empirical optimisation of the models. The Data Scientists can focus their efforts on the business needs.
Interpretability of the models: Easier interactions with business and marketing teams. Conformity with laws concerning privacy and personal data.