Prediction of Airbnb Rental Prices using Machine Learning (Results)

Context and Methodology:
This dataset was created as part of a machine learning project for predicting Airbnb rental prices in the USA (year 2023). It is used to evaluate models (in particular a RandomForestRegressor) on training and test data that were previously cleaned and processed from a large raw dataset (AB_US_2023.csv). The training data were versioned in DBRepo and loaded into the Modeling_Regression.ipynb notebook via the API. A 10-fold cross-validation tuning on the training data optimizes the model’s three hyperparameters.

Technical Details:
The dataset is split into a training set and a test set. All output files (model pickle, metrics CSV, plots) are stored in the results/ folder at the repository root. Additional configuration and sample data can be found in data/sampled_data/. The notebook retrieves only the current training and test PIDs via the DBRepo API, so no local CSVs are versioned in the repo.

Further Notes:
A detailed setup guide (downloading large CSVs, adjusting paths) and all other preprocessing notebooks are documented in the README. License: CC-BY-4.0.