MktPraxis
Interactive Textbook Module Syllabus

Neural Network Basics for Marketers

Every marketing platform now sells 'AI-powered' predictions, and many of them are neural networks. Learn what a network actually computes (a stack of tiny logistic regressions), how it learns by walking downhill, how to tell when it has memorized a small customer file and stop it in time, how to test honestly whether it beats a regression you could explain, what permutation importance and partial dependence do and do not tell you, and why no model should be trusted beyond the data it saw.

⏱️ ~65 min reading & interactives 📑 14 sections 🎓 Level: Advanced undergraduate 📐 Math: Formulas shown 📚 Courses: Predictive Analytics & ML, Marketing Analytics 🏷️ Topics: Machine learning & prediction, AI in research & analytics 🔗 Prerequisite: Prediction Fundamentals
Full Interactive Chapter (Account Required) The complete chapter reading, embedded live mini-widgets, inline Check-Your-Understanding questions, and progress tracking are available to enrolled students and verified instructors.
Open in Textbook Reader (Sign In) Request Free Instructor Preview

Chapter Section Outline (14 Sections)

Topics and managerial frameworks covered in Neural Network Basics for Marketers:

  1. 1. Why Every Marketer Gets Pitched a Neural Network
  2. 2. One Unit: A Weighted Sum and a Squash
  3. 3. Layers: Why Stacking Units Bends the Line
  4. 4. How a Network Learns: Walking Downhill in Fog
  5. 5. The Validation Curve and When to Stop
  6. 6. Why Networks Are Data-Hungry
  7. 7. Is the Network Even Worth It?
  8. 8. When the Network Wins, Ask Why
  9. 9. Opening the Black Box
  10. 10. Off the Edge of the Data: Extrapolation
  11. 11. Where Networks Earn Their Keep
  12. 12. Case Application — Kinoscope's Premium Push
  13. 13. Reference Card
  14. 14. What This Sets Up

Interactive Concept Mini-Widgets Embedded in This Chapter (2)

Students manipulate these interactive visual simulations directly inside the chapter reading before answering gated check-in questions:

Build a Bend: One Unit

regression; two hidden units summed together can draw a U that one cannot.

How Big a Step?

with failure modes on both sides.

Paired Browser-Based Analytics Calculators (2)

This chapter connects directly to these full MktPraxis analytical apps (with pre-loaded business case scenarios, CSV upload, and reproducible Python export):

🧠 Neural Network Playground

Build and visualize neural networks to solve complex marketing classification problems.

🕸️ Neural Network Lab

Train a neural network on your customer file, watch it learn, and find out on customers it never saw whether it beats a regression you can explain.

Pre-loaded Business Scenarios: Is the AI Worth It for the Save Offer? Ridgeline Roasters · Which 40,000 Get the Catalog? Copperline Outfitters · A Big Network on a Small Table: Tallyhop · Who Should See Premium 4K First? Kinoscope · Lift the Cap to $60k? Wrenfield Mattress