05 · Applied predictive modeling

Data-Driven Pricing Predictions for U.S. Manufactured Homes

A 2021-2023 manufactured-home price study comparing regularized regression and random forests on an out-of-sample 2023 holdout.

pricing prediction2021-2023

data → model → 2023 holdout

Role

Academic research - data preparation, predictive-model comparison, validation design, and written interpretation.

Context

Manufactured-home pricing depends on physical attributes and construction choices; the study frames the task as a reproducible supervised prediction problem for buyer comparison.

Tools

R · Lasso · Ridge · Random Forest · Cross-validation · HUD public data

Project overview

The paper evaluates whether home attributes can support more disciplined price estimates for manufactured homes using a nationwide public-use survey dataset.

Research question

How accurately can home characteristics such as size, bedrooms, foundation type, elevation method, and securing method explain 2023 manufactured-home prices?

Workflow

01Use HUD Manufactured Housing Survey public-use data spanning 2021 through 2023.
02Address missing values according to the report’s documented preprocessing choices and inspect non-linear size effects.
03Compare Lasso, Ridge, and Random Forest models with ten-fold cross-validation for model selection.
04Train on 2021-2022 data and evaluate on 2023 as an explicit out-of-sample holdout.

Data & methodology

Data sources

HUD Manufactured Housing Survey Public Use File
U.S. Census Bureau administered survey
2021-2023 manufactured-home records

Methods

OLS exploratory analysis
Cubic size terms and interactions
Lasso and Ridge regularization
Ten-fold cross-validation
200-tree Random Forest

Evidence boundary

Input→Preparation→Analysis→Output

The showcase is grounded in the local research paper and HTML output; the underlying public-use data is not copied into the portfolio.

Results & conclusion

Key outputs

Seven-page research paper
Out-of-sample R² comparison
Feature and model specification
Written interpretation of practical use

Documented findings

Random Forest produced the highest reported 2023 out-of-sample R² at 0.4967.
Lasso and Ridge reported 2023 holdout R² values of 0.4874 and 0.4833, respectively.

Conclusion

The documented models explain roughly half of the observed 2023 price variation using the available housing attributes, with similar cross-validation and holdout results.

Decision relevance

Attribute-based estimates can support structured comparison of manufactured-home configurations.
The remaining unexplained variation highlights the value of local costs, upgrades, and manufacturer context in real decisions.

Limitations

The model cannot observe every upgrade, local installation cost, or manufacturer difference.
Results are specific to the selected survey years and feature set.
The study is an analytical comparison tool, not a pricing recommendation.

Next project showcase

01 · Time-series machine learning

Predicting Bitcoin Price Direction

View all research & project showcases