Machine Learning and Design Pattern Best Practices

I recently wrote article on machine learning and design pattern best practices and published it on Linked In. It was inspired by the 43 rules that google published when creating machine learning enhanced web applications. Google split the rules into three phases:

  1. Pipeline, which looks at the infrastructure. This is really about tooling up your systems to be able to manage the influx of data and I call out two rules for further consideration.
  2. The second phase concerns Feature Engineering where we start to get a feel for the data and what it might be telling us.
  3. The third section on Growth, Refinement and Complex Models is by far the most interesting and sparsely populated list of rules from google.

After the low hanging fruit has been harvested in the second set of rules, it is the third that enables developers, analysts and architects to improvise and consider areas, ideas and techniques that have not been tried before. It is this area then that offers the most growth potential and is the most exciting for the future!

https://www.linkedin.com/pulse/best-practices-machine-learning-design-patterns-john-howard/