[FAQ Bot] NEW: Question 6: Using seed 9, combine train+validation, fill missing values

_questions/machine-learning-zoomcamp/module-2/033_b271d57642_mlq6-seed9-combine-fill0-train-r-0-001-rmse.md

@@ -0,0 +1,16 @@
+---
+id: b271d57642
+question: 'Question 6: Using seed 9, combine train+validation, fill missing values
+  with 0, train with r=0.001 — what is the test RMSE?'
+sort_order: 33
+---
+
+To get the RMSE for Question 6, you need to follow the exact homework procedure in one consistent pipeline:
+
+- Re-split the data using `seed = 9`.
+- Combine the training and validation parts back together (so the model trains on `train + validation`).
+- Fill missing feature values with `0` (using the constant fill specified by the question).
+- Train the regression model using `r = 0.001`.
+- Compute RMSE on the provided test dataset using the model’s predictions.
+
+If you still get a value that doesn’t match the multiple-choice options, double-check that you fully reinitialize the index/iteration variables used in the loop for question 5/6 (the homework notes that you should reinitialize like `Idx = np.arange(n)` for each question/`r` iteration), since stale state can shift results.