There’s a lot of excitement and growing interest in machine learning, even in mainstream business practice. The Data Scientist is a Star whose skills are much in demand. But for machine learning to be a practical success in real world, production settings, there’s a lot beyond the algorithm and model that must be done correctly. Is the question or issue being addressed by ML appropriate for the specific goal? Can models be evaluated in a meaningful way and deployed and maintained in production with the expertise available? Is there sufficient domain knowledge to make certain the project is addressing the right issues? Is there a way to take practical action on the insights gained through machine-driven decisions?
This presentation will explore real world examples to discover some of the pitfalls to avoid and to show a variety of tips and best practices - from planning to model management and deployment in production -- that can help make your machine learning systems successful.