Graded Quiz: Predictive Analysisis:Applied Data Science Capstone (Applied Data Science Specialization) Answers 2025
1. Question 1
How many records were there in the test sample?
-
✅ 18
(Enter: 18)
Explanation: The dataset of 90 Falcon 9 launches is split using an 80/20 rule in the lab:
90 × 0.2 = 18 test records.
2. Question 2
For Support Vector Machines, which kernel has the best result on the validation dataset?
-
❌ rbf
-
✅ linear
-
❌ sigmoid
Explanation: In the ML lab, the linear SVM kernel gives the highest validation accuracy compared to RBF and sigmoid.
3. Question 3
After selecting the best hyperparameters for the Decision Tree classifier, what was the accuracy on the test data?
-
❌ 83.33%
-
❌ 73.33%
-
✅ 93.33%
Explanation: Running the final tuned Decision Tree model in the lab yields a 93.33% test accuracy on the first run.
🧾 Summary Table
| Q# | Correct Answer | Key Concept |
|---|---|---|
| 1 | 18 | Test set = 20% of 90 records |
| 2 | linear kernel | SVM linear performs best on validation set |
| 3 | 93.33% | Final tuned Decision Tree accuracy |