Our normalization validation exercises encompass several tests, each with a specific focus on the fidelity of our footfall data for ranking and time series analysis. By applying these tests, we aim to affirm that our normalization method not only provides a static snapshot of location footfall but is also capable of tracking and representing the dynamic nature of footfall activity over time. This temporal analysis is key for applications that require understanding and predicting patterns of movement and occupancy in specific areas.

1. UK museum ranking by visit volume (monthly average)

The first test examines whether our normalized data can accurately reproduce the actual rankings of UK museums based on their visitation volumes, averaged monthly. This test is crucial to ensure that our normalization process can reliably reflect the relative popularity of locations based on visitor activity.

Ranking correlation score of 0.9 (Spearman’s rank coefficient)

2. Month-to-month visit matching of each UK museum

In the second test, we perform a detailed month-to-month visit comparison for each UK museum. This step is intended to verify that our normalization is consistent over different time periods and that it accurately reflects the actual visits, ensuring that the process is robust across time without introducing temporal distortions.

3. Week-by-week Liverpool BID area football count

The third test is focused on the Liverpool Business Improvement District (BID), where we analyze the weekly footfall counts to evaluate the performance of our normalization in time series analysis. This allows us to assess whether our normalized counts can maintain temporal integrity and accurately reflect footfall trends over time within the same location. It tests the ability of the normalization process to capture and represent the nuanced changes in visitor patterns that occur from week to week.

Last updated