Building Predictive Models in ArcGIS Pro
Forecasting and prediction models are a cornerstone of our work as aspiring policy and data analysts – this tutorial did a great job highlighting what happens behind the scenes when we build our models, and I especially appreciated that it encouraged exploration to fully understand what happens when we manipulate our setup.
I can see similar predictive models being useful in developing at-risk areas that might require more targeted social services – a contemporary example might be a state trying to predict which counties are at greater risk for COVID death and infection. However, I can see this expanding beyond COVID analysis to a host of other social and economic indicators, such as rates of home ownership, risk of defaulting on debt, HIV infection rates for certain communities, and so on to create more focused policies and programs. I was curious about what other prediction tools or algorithms are available through GIS – in classes like ManSci, we focus on exponential smoothing and seasonal forecasting over things like random forests because we work with relatively small datasets with distinct patterns. I would be interested in continuing to poke around and work with other prediction tools to see if there are other approaches we can pick for differently distributed data!