3 ms·

Is Starbucks good at predicting gentrification? Presumably there is some criteria someone is using to gauge the feasibility of a location vs others. If you ca

• financltravsty 2y ago
  Look at demographic data & data that correlates with it. The Census and related organizations offer very fine-grained data (both in characteristics/factors & in level of specificity of location). In the past, I had built a data brokerage offering clean and centralized data in this area -- the equivalent of Netflix for data catalogs (back when you could say this with a straight face). It is not a hard problem. You can (and many did) throw entire datasets at a random forest as a first-pass, and then retrain with only the features that pass a significance threshold. Algorithmic identification of real estate investment opportunities for the purpose of corporate expansion is not a new thing. Most serious shops do it. The techniques are simple, but it's the prep and decision-making after that are difficult.