Machine Learning: Optimization with Random Exploration
This project focuses on evaluating how random exploration can be used for quick and practical optimization of a machine-learning model designed to predict whether a car will be sold based on variables such as mileage, model year, and price. The project utilizes RandomizedSearchCV to test random values within a specified range for the hyperparameters max_depth, min_samples_split, min_samples_leaf, and criterion of a DecisionTreeClassifier model. Additionally, the project explores the use of GridSearchCV and cross-validation to optimize the model while balancing computational resource consumption.
The goal of this project is to demonstrate how random exploration can be an efficient method for hyperparameter optimization in machine learning models. By comparing exhaustive and random searches, the project aims to find the best model with optimized hyperparameters without overusing computational resources.
The project focuses on the following:
- Explore parameter spaces randomly.
- Use RandomizedSearchCV for efficient search optimization.
- Compare exhaustive search with random search.
- Perform optimization with and without cross-validation when applicable.