Guided automation for machine learning

Paolo Tamagnini, Simon Schmid, and Christian Dietz are data scientists and software engineers at KNIME.

We propose here a blueprint solution for the automation of the machine learning lifecycle.

The price to pay for automated machine learning (aka AutoML) is the loss of control to a black box kind of model. While such a price might be acceptable for circumscribed data science problems on well-defined domains, it might prove a limitation for more complex problems on a wider variety of domains. In these cases, a certain amount of interaction with the end users is actually desirable. This softer approach to machine learning automation — the approach we take at KNIME — is obtained via guided automation, a special instance of guided analytics.

[ The essentials from InfoWorld: The 5 best programming languages for AI development. • Why you should use Python for machine learning. • Julia vs. Python: Julia language rises for data science • 5 essential Python tools for data science—now improved • Get started with Anaconda, the Python distribution for data science. • What’s new in the Anaconda distribution for Python. | Keep up with hot topics in programming with InfoWorld’s App Dev Report newsletter. ]

As easy as the final application might look to the end user, the system running in the background can be quite complex and therefore not easy to create completely from scratch. To help you with this process, we created a blueprint of an interactive application for the automatic creation and training of machine learning classification models.