This repository contains Python code for building and comparing linear regression models using the Iris dataset. Two models are implemented to predict the petal width based on different independent variables (sepal length and petal length). The models are built using the probabilistic programming library PyMC.
- Data Source: The Iris dataset is imported from the "iris.csv" file using the Pandas library.
- Correlation Analysis: Pearson's correlation coefficients are calculated for pairs of features to decide the independent variable for each model.
- Model 1: Linear regression model with petal length as the independent variable.
- Model 2: Linear regression model with sepal length as the independent variable.
- R-Squared Comparison: R-squared values are calculated to evaluate the goodness of fit for each model.
Make sure you have the following dependencies installed:
- Python (>=3.6)
- NumPy
- Pandas
- PyMC
- Arviz
- Matplotlib
Install the required dependencies using:
pip install numpy pandas pymc arviz matplotlib- Clone the repository:
git clone https://github.com/Moeez-Muslim/Linear-Regression-Models-for-Iris-Dataset.git- Navigate to the project directory:
cd Linear-Regression-Models-for-Iris-Dataset- Run the Jupyter Notebook:
jupyter notebook linear_regression_iris.ipynb- Explore the analysis and results in the notebook.
The R-squared values suggest that Model 1 (using petal length) provides a better fit to the data compared to Model 2 (using sepal length).
- Moeez Muslim [email protected]
This project is licensed under the MIT License - see the LICENSE file for details.