This project demonstrates the end-to-end workflow of training a linear regression model, saving the trained model and datasets, and deploying a Streamlit web application for predictions and visualizations. The goal is to provide an example of a basic machine learning pipeline using Python and popular libraries.

The project is divided into two main parts:

1. Model Training:

   • Train a linear regression model using synthetic regression data.
   • Save the trained model and datasets for future use.
   • Evaluate the model's performance with metrics like Mean Squared Error (MSE).

2. Streamlit App:

   • Provide a user-friendly interface for making predictions using the saved regression model.
   • Visualize the relationship between input features, actual target values, and predictions.

• Train and evaluate a linear regression model using scikit-learn.
• Save the trained model and datasets using joblib.
• Interactive prediction using a Streamlit app.
• Visualization of differences between actual and predicted target values.

The following Python libraries are required:

• numpy
• scikit-learn
• joblib
• streamlit
• matplotlib

Install all dependencies using the provided requirements.txt file:

pip install -r requirements.txt

git clone mehalyna/AI_Fund_final_task
cd AI_Fund_final_task

Run the train_and_save_model.py script to train the regression model, evaluate it, and save the necessary files:

python train_and_save_model.py

This script will:

• Train a linear regression model.
• Save the model as linear_regression_model.joblib.
• Save the datasets (X.joblib and y.joblib).

Run the Streamlit app for making predictions and visualizing results:

streamlit run streamlit_app.py

• Open the URL displayed in the terminal (usually http://localhost:8501).
• Adjust the slider to select a feature value between -3.0 and 3.0.
• Click the "Predict value" button to:
  • See the predicted target value.
  • Visualize the difference between the actual and predicted values.

├── train_and_save_model.py  # Script for training the regression model and saving datasets.
├── streamlit_app.py         # Streamlit application for prediction and visualization.
├── requirements.txt         # Required libraries for the project.
├── README.md                # Project documentation.
├── linear_regression_model.joblib  # Saved regression model.
├── X.joblib                 # Saved feature dataset.
├── y.joblib                 # Saved target dataset.

1. Training Output Example:

   Model trained and saved as linear_regression_model.joblib.
   Datasets saved as X.joblib and y.joblib.
   Mean Squared Error: 416.80
   

2. Streamlit App Interface:

   • Input slider to select feature values.
   • Button to predict and visualize results.
   • Scatter plot showing dataset, actual, and predicted values.

• Ensure that linear_regression_model.joblib, X.joblib, and y.joblib are in the same directory as the Streamlit app script for proper functionality.
• Modify the slider range or visualization options in streamlit_app.py to suit your dataset or use case.

This project is licensed under the MIT License. Feel free to modify and use it for educational purposes.