This document discusses predicting house sale prices using regression algorithms. It analyzes a dataset of 1460 houses with 81 variables like size, age, and sale price. Missing lot size (LotFrontage) values are filled in using neighborhood averages. Highly correlated variables like size metrics are identified. Models like linear regression, KNN, random forest, and decision trees are trained and tested on scaled data, achieving test R-squared scores from 0.71 to 0.85. Linear regression and random forest performed best.