| Customer | Monthly Spend ($) | Support Calls | Contract Length | Churned? |
|---|---|---|---|---|
| 1 | 50 | 5 | 3 | Yes |
| 2 | 120 | 1 | 24 | No |
| 3 | 80 | 3 | 12 | No |
| 4 | 35 | 8 | 2 | Yes |
| 5 | 200 | 0 | 36 | No |
Apply sigmoid function: P(Churn) = 1/(1+e^(-z)). Calculate z = w₀ + w₁·spend + w₂·calls + w₃·contract. Use gradient descent to optimize weights.
Understanding the mathematical foundation helps apply the algorithm effectively in real-world scenarios.
Use the method learned in this course to solve a similar problem with your own dataset.
Explain the mathematical principles behind Logistic Regression & Classification and when to use it vs alternatives.
Review the mathematical derivations in the main course and understand the assumptions, strengths, and limitations of this approach.