Architecting the Industrial Internet

Machine learning tools

Machine learning tools provide a development environment for experimenting with mathematical modeling algorithm alternatives and then selecting the most appropriate model. Most often, experimentation initially occurs using a sample of the full data set.

Once the model is defined, many such tools also automate the process of building application code. Code generation into languages such as C#, Python, and R is typical. The code can then be deployed to the appropriate component in the architecture.

The tools usually come packaged with various statistical functions, machine learning algorithms, and text analytics functions and feature the ability to insert custom code. Statistical functions used to measure the variation ...