Life, Death, and Shopping
summary
Claims a practical pipeline that turns past event logs into labels for near‑future purchase prediction using sliding run dates and aggregated temporal features, not theory. Uses moving windows and PySpark window functions (SQL/pandas equivalents discussed), builds expanding vs rolling aggregates and TF‑IDF weighting plus sequence (anagram) methods for funnels. Shows an end‑to‑end feature engineering pattern on retail clickstreams and mentions scale, but provides no concrete accuracy or business impact numbers.
AI-written summary · about
chapters (6)
- 0:00Cell death introduction
- 1:25Survival analysis for prediction
- 2:34Predicting purchases with windows
- 6:37Aggregating activity with windows
- 9:30Feature engineering for behavior
- 11:18Summary
related talks
27mLife, Death, and Shopping
38mOptimizing Retail Forecasting Through AI Models
36mThe Future of Traditional Shopping Driven by Customer Centric Approaches by Santiago Cabrera Naranjo
07mTransforming Big Data into Real-Time Product Insights
23mDelivering Retail Analytics & Data Projects With Data Created For Purpose
41mFast Data - an Overview of the Concepts, Architecture and Technology of Streaming Analytics
42mBuilding a Data Driven Organization
23mArtificial Intelligence for Store Assistance