Surpise Housing Assignment

An Advnace regression model using Rigde and Lasso regression to predict the sales price of housing in Australia.

Table of Contents

• General Info
• Technologies Used
• Conclusions

General Information

Conducted Analysis with the following technique

• Data Cleaning
• Data Manipulation
• Univariate Analysis
• Bivariate Analysis
• Data Visualization
• Dummy Variable creation
• Ridge Regression
• Lasso Regression

Conclusions

• Optimal value of lambda for Ridge Regression = 10
• Optimal value of lambda for Lasso Regression = 7.0
• The variables significant in predicting the price of a house are :-
  • GrLivArea
  • OverallQual_9
  • OverallCond_9
  • OverallQual_8
  • Neighborhood_Crawfor
  • Functional_Typ
  • Exterior1st_BrkFace
  • SaleCondition_Alloca
  • CentralAir_Y
  • TotalBsmtSF
  • Neighborhood_Somerst
  • TotalBsmtSF
  • Condition1_Norm
• Lasso performs feature elimination

Technologies Used

• Python
• Pandas
• Matplotlib
• Seaborn
• Jupyter Notebook

Contact

Created by @prasoonmhwr - feel free to contact me!