Missing Data Strategies

Missing data is everywhere in the real world. Sensors fail, forms go half-filled, databases merge badly. The easy fix — dropping every incomplete row — throws away information you might need. Filling blanks with the column mean sounds reasonable, but what if the data is skewed? In this quest you'll create three copies of a housing dataset, apply a different missing-data strategy to each, train a LinearRegression model on every version, and compare their R-squared scores. The numbers will speak for themselves.