SQL Server Predictive Analytics: Customer Stories
summary
Shows a full SQL Server workflow: Analysis Services models built from relational data via an Excel add‑in, 30% held‑out validation, and a profit chart that finds a 35.69% threshold giving a max profit (~$63,875) for a bike‑mailing campaign; the chosen decision‑tree model is deployed as a Reporting Services URL that refreshes daily. Real customers include an HMO with 3.7M members, 4.7 TB warehouse using nested tables (10 nested inputs) to score risk and intervene, and EB Games (400 stores, 1.1M customers) using association rules, clustering and a time‑series ARIMA combo for per‑store forecasts.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:05Predictive analytics in BI
- 7:27Excel demo: predictive marketing
- 17:37Evaluating predictive models
- 27:37Healthcare success stories
- 35:45EB Games retail forecasting
- 42:10Key takeaways
- 43:13Resources for SQL Server
- 50:33Q&A
related talks
56mPredictive Analytics for Practitioners
01h 05mAzure Machine Learning for the Developer
45mRebooting your SME Data Analysis Strategy
27mEnterprise AI at Scale
27mEnterprise AI at Scale: Success Stories and Insights
18mEmpowering Business Through Data
53mMachine Learning in E‑commerce
01mDatabricks Customers