**Our purpose is to identify/delineate agglomeratons within a CDP
**We use lowest-highest layers, which correspond to unique hull counts with 'stable' hulls, in a regression at the CDP level, with a performance measure (next year's venture investment) as the dependent variable.
===Suggestions and Questions===
Terminology - Match to audience but accurate.
*Propose: Hulls -> Clusters. Issues: mixes statistical def, also what about boundary? Alternative: Area.
*Census place -> cities.
*Lowest-highest -> representative
*Understand regression to select hull count
*Heat map!
=Ed's Big Effort=