R/F Taskforce Meeting 20220927
Agenda
Status update on testing and models
Model 2 is now in the QA environment
Model v2 is performing very well and has an RSME of 4.9 for all user test packages from the insights suite. This is compared to 18.2 for production and 8.2 for staging (v0). The lower the RSME the better.
Use case examples and comparisons
Comparing Production, Staging, and QA Reach %s
Documentation Update
Geopath to refine further, then will utilize sub-group for approval before seeking approval from larger group
Â
Overall Model Comparison
Â
Â
Â
Test Case Comparisons
Example 1
Chicago DMA, Persons 18+, 4 weeks
Â
Production Reach | Staging Reach | QA Reach | Observed Reach | TRPs | Frequency | *Notes |
22.49 | 27.45 | 27.81 | 32.60* | 245.68 | 8.83 | Observed is against PopFacts 18+ |
Â
Example 2
Washington DC DMA, Persons 21+, 2 weeks
Production Reach | Staging Reach | QA Reach | TRPs | Frequency |
15.46 | 15.50 | 21.26 | 73.74 | 3.47 |
Â
Â
Â
Example 3
Chicago CBSA, Persons 18+, 4 weeks
Â
Production Reach | Staging Reach | QA Reach | TRPs | Frequency |
2.71 | 8.11 | 19.38 | 57.25 | 2.95 |
Â
Â
Example 4
Los Angeles DMA, Persons 18+, 4 weeks
Production Reach | Staging Reach | QA Reach |
28.55 | 48.40 | 45.42 |
Â
Â
Example 5
Atlanta DMA, Males 25-54, 4 weeks
Production Reach | Staging Reach | QA Reach | TRPs | Frequency |
18.33 | 46.71 | 39.65 | 367.82 | 9.28 |
Â