Machine Learning Application

This is a Python-based machine learning application built using Streamlit. It demonstrates the implementation of various supervised machine learning algorithms for classification and regression tasks.

Introduction

Machine Learning is concerned with using algorithms that automatically improve through iteration to produce a model that can predict outcomes of new data. This application allows users to explore different datasets, choose algorithms, and evaluate their performance.

Features

• Data Wrangling: The application provides options to select and visualize different datasets, such as Iris, Breast Cancer, Wine, California Housing, Boston House, and Diabetes.
• ML Algorithms: Users can choose between supervised and unsupervised learning. In the supervised learning section, they can select between classification and regression tasks and further specify the algorithm to use.
• Classification: For classification tasks, algorithms like K-Nearest Neighbors (KNN), Support Vector Machines (SVM), and Decision Trees are available. The application allows users to tune hyperparameters and evaluates the model's accuracy.
• Regression: For regression tasks, algorithms like Linear Regression, Support Vector Machines (SVM), K-Nearest Neighbors (KNN), and Decision Trees are available. Users can tune hyperparameters and evaluate the model's performance using Mean Absolute Percentage Error (MAPE).

Installation and Usage

1. Clone the repository:

git clone https://github.com/your-username/machine-learning-app.git

2. Install the required dependencies:

pip install -r requirements.txt

3. Run application :

streamlit run machine_learning_app.py

4. Open the application in your web browser:

Local URL: http://localhost:8501
Network URL: http://192.168.X.X:8501

5. Interact with the application by selecting different datasets, algorithms, and hyperparameters to train and evaluate machine learning models.

Dependencies

• streamlit
• pandas
• numpy
• scikit-learn
• matplotlib
• seaborn

File Structure

The code is organized into the following files:

• machine_learning_app.py: The main file that runs the machine learning application and handles user interactions.
• data_loader.py: Contains the DataLoader class for loading and preprocessing datasets.
• model_trainer.py: Contains the ModelTrainer class for training and evaluating machine learning models.

License

This project is licensed under the MIT License.

Feel free to explore and use this machine learning application to gain insights into different datasets and algorithms!

==>

this project was collaboriation with 3 students as a graduation project it was has alot more but it deleted because the dependeinces was not stable