Skip to content
Bosaj edited this page Aug 23, 2026 · 1 revision

FAQ

Why only three features (TotalSqFt, BedroomAbvGr, FullBath)? This task is intentionally scoped to a simple linear regression exercise — the Prodigy InfoTech internship brief asks specifically for square footage and bed/bath counts, not the full ~80-column Ames dataset.

Is an RMSE of ~46,378 good? It's a reasonable baseline for a three-feature linear model on a dataset where sale prices range roughly from $35,000 to $750,000+. Kaggle leaderboard solutions using the full feature set and gradient-boosted models achieve substantially lower RMSE (often under $15,000), since they capture location, quality, and condition effects this notebook doesn't model.

Where do train.csv and test.csv come from? They're the standard files from Kaggle's House Prices — Advanced Regression Techniques competition and must be placed in data/ before running the notebook.

Can I submit data/submission.csv to the Kaggle competition? Yes, it's formatted for that competition's submission format (Id, SalePrice), though its score reflects the simplified three-feature model, not a competition-tuned pipeline.

Does CI check the RMSE value? No — CI validates notebook structure and dependency installability only. It doesn't execute the training cells.

Clone this wiki locally