Machine Learning

When making buy-sell-hold decisions, investors need timely assessment of the in-place conditions of both existing and potential investment sites. Linnaean’s location intelligence toolkit integrates high-resolution structural, demographic, and economic information with travel-time accessibility analytics to provide granular comparisons starting from the address level.

Proprietary data and metrics tailored to a firm’s differentiated investment thesis can be integrated into these overviews, combining conventional demographic or economic target criteria with underwriting information specific to an investment committee’s goals.

In addition to expertise in Census and similar survey information, Linnaean has extensive work experience and partnerships with near real-time alternative datasets such as GPS foot traffic monitoring and credit card transactions.

Utilizing these datasets on recent trends in conjunction with high-granularity estimates of in-place demographics and economics allows inference of hard-to-measure but critical characteristics such as the buying power commuting past a retail location. Linnaean can also produce “nowcasts” of the vitality of locations, including resilience in economic shocks such as the COVID pandemic.

Responding to the stakeholders’ desire for more timely and granular information on local demographics and economics, Linnaean has developed and deployed multiple methods of algorithmically imputing demographic information like educational attainment and household incomes at geographic levels more specific or more timely than the officially-reported aggregations.

These indicators are then used in visualization applications to allow stakeholders to directly view information around their city-block level areas of interest; and as input features to models used to predict future performance and risk.

Firms that engage Linnaean are without exception experts in their application fields, but frequently making their first investment into advanced analytics. Translating model results into actionable, relatable information is just as crucial as getting the prediction right in the first place.

Interactive exploratory software, sensitivity testing and “explainable AI” prediction attribution tools are therefore core to Linnaean’s approach at every stage of collaboration.