Open Source Model Infrastructure

What does Open Source Model Infrastructure Mean?

The term “open source” refers to a philosophy of sharing code, ideas, and innovations, enabling the knowledge base of an entire industry to grow more quickly than if that knowledge remained proprietary.

In an open source model infrastructure, some or all of a business or technology is built upon open source principles. It is often thought to be primarily linked to software, but there are open source communities for hardware, robotics, manufacturing, services, economics — even an open source eyewear brand.

There is also a big open source community that supports machine learning and artificial intelligence (AI). Two of the most popular open source frameworks are the Python scikit-learn and a number of R machine learning packages. There are many others, ranging in size from the Tensorflow library to countless individual and group contributors.

Why is Open Source Model Infrastructure Important?

At the heart of machine learning technology lies algorithms developed by the open source community that help analysts uncover patterns and insights from historical data. The community includes academic institutions; companies such as Google, Microsoft, and Yahoo; and individual developers. Since the community values openness and collaboration, everyone gets the opportunity to use these algorithms, verify and discover what works well and what doesn’t, and iterate on the algorithms to continuously make new and better ones.

DataRobot + Open Source Model Infrastructure

The open source model infrastructure is an essential underpinning of the DataRobot AI platform, which uses open-source algorithms for most of its models because they are some of the best available. The algorithms contained in our model blueprints – automatic combinations of data preprocessing steps and machine learning algorithms – come from a breadth of open source software frameworks, including software available in programming languages like Python and R and libraries such as Tensorflow, XGBoost, DMTK, and Vowpal Wabbit.

See our robust selection of open-source and proprietary algorithms