SketchNotes on the Data Science of Beer Styles
Dramatic success in machine learning has led to a torrent of Artificial Intelligence (AI) applications. Continued advances promise to produce systems that will perceive, learn, decide, and act on their own. However, the effectiveness of these systems is limited by the machine’s current inability to explain their decisions and actions to human users. For instance, a mobile app might recommend a beer of a particular style to a user and the user might not know why. Visualizing the recommended beer in the context of other beers and beer styles is one way an AI agent can explain itself. Seeing what an AI was thinking can help build trust. Here are some sample visualizations of beer tasting notes (pumpkin, floral, citrus, grapefruit & dark) by beer style.
Coverage:
Catherine Madden (@catmule) sketched notes from Dr. Kaufhold’s “Visualizing Beer Styles” talk at the MLK library in Washington DC.
Beer analytics site BeerGraphs discussed Dr. Kaufhold’s embedding methods for beer review data.
Beer Heat Maps!