Cook Part 3

Part A (10%)

Begin to finalize your report by only including report/website quality output. By this, only include stylized output (do not use base R print). Include your code in your report using code_folding: hide. Avoid any package loading or other incidental inclusion of output in your report. This could mean using stargazer to show regressions, DT::datatable to present tables, and adding titles/labels to plots. Give all plots and tables appropriate captions. Write two to three sentence introductions for the different sections of your report. Your report should introduce property assessment in Cook County, Illinois (as defined in part 2A), your assessment prediction model (from 2C), and a conclusion. Note that Part 2B will not be continued into this assignment.

Part B (40%)

Feature engineering. You have now created a base models and evaluation metrics from last week. Investigate creating at least two new predictors and analyze if they improve your model(s). Some possibilities:

• Census variables (such as income, race)
• Sale price per square foot for neighborhood
• Geographic areas (tract, township, etc.)
• Distance to geographic features (CTA, Lake Michigan, Nearby Counts of Schools/PINs/Transit/etc)
• Time adjustments

Specifically compare model performance between the base model and the new predictors. If these features do not show sufficient improvement, state that and then remove them from your model.

Part C (50%)

Prediction. Create “out of sample” predictions or “assess” all properties. By this, predict assessment/valuation for homes which did not sell (e.g. predict/assess properties for 2023 from part 2C). Note if you are having trouble with this step that you cannot use any information specific to the sale of a property for out-of-sample prediction. In other words, we use information on sale prices to determine the true value of homes but we do not know this information for homes which did not sell and if we want to make a prediction for these homes we cannot use sale price information as a predictor.

A helpful data framework for this section would be to create a dataset of all properties. You would then label all the properties which sold (leaving un-sold properties unlabeled) and create your testing and training data by filtering only to properties which were labeled. After training, you can then augment your full dataset of labeled/unlabeled data to get out of sample predictions. This framework is helpful because we need the data used to make predictions to have exactly the same format as the data used to train the model.

Using at least one table, plot, and map, present aggregated information on your predictions. For example:

• a table showing the accuracy of your assessments compared to the assessor’s assessments by region or property value
• a plot showing the distribution of predicted market values
• a map which shows summary information of market values/assessment accuracy across your township