Question 1: Short answer
mediumAI-checkedClassificationSix training SMS are encoded by whether four words occur:
| message | block | free | prize | urgent | label |
|---|---|---|---|---|---|
| 1 | 0 | 0 | 1 | 1 | spam |
| 2 | 0 | 1 | 0 | 1 | ham |
| 3 | 1 | 0 | 0 | 0 | ham |
| 4 | 0 | 1 | 1 | 1 | spam |
| 5 | 0 | 1 | 0 | 0 | spam |
| 6 | 1 | 0 | 0 | 0 | ham |
A Bernoulli naive Bayes model is fit by plain counting (no smoothing). The new message "URGENT! Free!!" has block = 0, free = 1, prize = 0, urgent = 1. What is after normalizing the two class scores?