在本地使用 AWS ML model 随机森林砍伐森林

Question

I wonder if it is possible to deploy Random Cut Forest (RCF) built-in algorithm of SageMaker to the local mode. I haven't come across any sample implementation about it. If not, can we simply say that models trained using RCF are limited to be consumed inside the platform via Inference Endpoints?

I got this error when I tried to do so.

Answer 1

indeed you're right, SageMaker Random Cut Forest cannot be trained and deployed locally. The 18 Amazon SageMaker Built-in algorithms are designed to be trained and deployed on Amazon SageMaker. There are 2 exceptions: SageMaker BlazingText and SageMaker XGBoost, which can be read with their open-source counterparts (fastText and XGBoost) and used for inference out of SageMaker (eg EC2, Lambda, on-prem or on your laptop - as long as you can install those libraries)

There is an open-source attempt to implement the Random Cut Forest here https://github.com/kLabUM/rrcf ; I don't think it has any connection to SageMaker RCF codebase so results, speed and scalability may differ.