How can machine learning be used to optimize control of a waste water treatment plant? This talk is a practical walk-though of an applied AI project. It discusses how to use a neural network trained in Tensorflow to replace failing sensors, and how to use reinforcement learning for optimizing real-time control of an industrial facility.
Robert van Straalen is a Lead Data Scientist at Data Science Lab, where he helps various organizations with setting up data science teams and workflows, as well as developing & deploying machine learning models.
This talk was presented during the Data Science Conference / 5.0 as a part of the Deep Learning & AI track.
