In this project, I use XGBoost, a powerful machine learning library, to predict real estate prices. The goal is to build a predictive model, evaluating the performance and tuning hyperparameters for optimal results.
Categorical columns are encoded to facilitate model training by converting them into a numerical format.
The project addresses missing values in the dataset through appropriate strategies to ensure data integrity.
The DMatrix object is used to efficiently handle the dataset, optimizing it for XGBoost.
The dataset is divided into training and testing sets to assess the model's performance accurately.
This section covers the configuration of model settings and parameters to achieve the desired outcome.
Mean Absolute Error (MAE) is computed to quantify the model's prediction accuracy.
The project explores hyperparameter tuning using the HYPEROPT library to enhance the model's performance.
K-fold Cross Validation is employed to assess the model's robustness and generalization across different subsets of the data.
An analysis of feature importance is conducted to understand the variables that significantly impact the model's predictions.