In this age of Artificial intelligence (AI) for everyone, much thought has been given to how to build and deploy machine learning solutions. This article details some of the work required both before and after the machine learning solution has been developed.
- Specifics about What data is required? Am I ready to develop an machine learning solution? Have I captured enough failure events to build a useful predictive maintenance solution?
- What will be the actual product delivered? Once I have a solution, how can I use this to improve my maintenance strategy?
- What should we do after we have brought these solutions into our business practice? How can I improve my solution going forward?
This article was written specifically for predictive maintenance settings when the goal of the model is to remove the events the model is predicting. However, many of these ideas are just as useful for any machine learning solution you are planning on bringing into your business practice.
Use intelligence when deploying an Artificial Intelligence solutions into your business practice.
(Created by a Microsoft employee)