This repository contains Python scripts for building custom machine learning pipelines for classification and regression tasks, as well as a Jupyter notebook demonstrating their usage with sample datasets.
This Python script provides a custom machine-learning pipeline for classification tasks. It initializes multiple classifiers, performs hyperparameter tuning, and conducts cross-validation for all models. The script then selects the best-performing model based on their accuracy.
Similar to the classification pipeline, this Python script focuses on regression tasks. It initializes multiple regressors, performs hyperparameter tuning, and conducts cross-validation for all models. The script selects the best-regressing model based on their mean squared error.
This Jupyter notebook demonstrates the usage of both the classification and regression pipelines. It uses the well-known Iris dataset for classification and the Boston housing dataset for regression. The notebook includes step-by-step examples, providing a clear guide on how to implement the pipelines, perform hyperparameter tuning, and assess model performance.
-
Clone the repository to your local machine:
git clone https://github.com/kipsangmarion/ml-pipelines.git
-
Navigate to the repository:
cd ml-pipelines
-
Run the
Example_Usage.ipynb
notebook in your Jupyter environment to see practical examples of using the classification and regression pipelines with sample datasets.
Feel free to customize the scripts and notebook according to your specific datasets and requirements.