Description
PROBLEM DEFINITION
Your goal is building a prediction model for real estate sale prices in Manhattan.
DATA
You have Manhattan real-estate sales data for real estate data between August
2012 – August 2013.
TASK
1. Clean the data and import to R.
2. Identify missing data and propose a method to handle the problem of
missing data.
3. Visualize 5 interesting trends.
4. Build a uni-variate linear regression model with only land square feet as
the input variable.
5. Build a multivariate linear regression model.
6. Design an experiment using multivariate linear regression model and kfolds cross validation.
7. Bonus (extra 20 points): Build a stepwise regression model and identify 5
top attributes in terms of information content.
OUTPUT
A zip file containing the following
• Cleaned data
• Report that contains:
o The data import method
o Visualizations
o Linear regression results summary
o Conclusions
o Bonus:
• All the source code
