Public Health Hotspot Allocator

Public health teams rarely have enough mobile response units — naloxone outreach vans, vaccination teams, disease-surveillance staff — to cover an entire region equally, so where they go matters. This simulation scatters synthetic incident data (overdoses, case reports) across a 3D city grid, runs a real DBSCAN density-clustering pass to find hotspots without assuming their number or shape in advance, and then compares two deployment strategies: spreading response units evenly across the map versus targeting them at the densest clusters. Adjust the DBSCAN radius and minimum-points threshold to watch clusters merge, split, or dissolve into noise, tune the unit count and response radius, and switch between allocation modes to see the mean response distance and coverage percentage shift live — a concrete, quantified case for density-based ML targeting over uniform coverage.