Using California as a case, we'll consider forest structure data from the California Forest Observatory, also generated using machine learning models. At 10 meters per pixel, this amounts to 400 million pixels per time slice (4) per variable (5). To concentrate on wildfire risk for households, we use the fire science concept of “defensible space” and buffer buildings to extract surrounding forest conditions. Then we apply supervised techniques such as Random Forests so that we can explore the factors correlated with particular outcomes.
To learn more about HEAVY.AI, visit http://www.heavy.ai.