This notebook includes coding and notes for predicting California house prices using multiple and polynomial regression in python.
-
Successfully trained, tested and deployed California house price machine learning model in python using multiple and polynomial regression algorithms.
-
Acheived an r2 score of 0.5912 using both multiple and polynomial regression on the highest performing features.
-
While the model is not as accurate as expected it compares favorably with an r2 score of 6.110 for the entire training set using the same features.
-
Visualized feature performance with correlation heatmaps and pairplots.
-
Created custom function to visualize selected dataframes, features, and plot layout.
-
Visualized results for all models.
-
Deployed the model on fictional housing data.