📈 Financial Prediction System

An interactive web application for financial prediction using machine learning with SVR (Support Vector Regression) and advanced visualizations.

🌟 Features

• Real-time Prediction: Predictions based on pre-trained SVR models
• Interactive User Interface: Intuitive and modern Streamlit interface
• Advanced Visualizations: Interactive 2D and 3D charts with Plotly
• Prediction History: Track and analyze all predictions made
• Analytical Dashboard: Complete overview with statistics and trends
• Multidimensional Analysis: Customizable 3D visualizations to explore relationships between variables

📋 Prerequisites

• Python 3.7 or higher
• pip (Python package manager)

🚀 Installation

1. Clone this repository or download the project files

2. Install the required dependencies:

pip install -r requirements.txt

💻 Usage

1. Ensure the model files are present in the project directory:

   • SVR.joblib: Pre-trained SVR model
   • scaler.joblib: Scaler for data normalization

2. Launch the Streamlit application:

streamlit run app_mod.py

3. The application will automatically open in your default browser

4. Use the sidebar to enter input parameters:

   • PrixArgent: Silver price
   • Day: Day (1-31)
   • month: Month (1-12)
   • year: Year (2000-2030)
   • Réserve extérieur: External reserves
   • Export: Export value
   • Import: Import value
   • Indice des prix à la consommation: Consumer Price Index (CPI)
   • revenu personnel: Personal income
   • Dépenses de consommation personnelle: Personal consumption expenditures

5. Click the "Prédire" (Predict) button to get a prediction

📊 Application Features

Main Dashboard

• Current prediction displayed in large format
• Bar chart of parameter values
• Pie chart showing value distribution

Trend Tab

• Visualization of prediction evolution over time
• Line chart with markers to track changes

3D Tab

• Interactive exploration of relationships between three variables
• Customizable axes (X, Y, Z)
• Color coding based on prediction values

Statistics Tab

• Histogram of prediction distribution
• Box plot for statistical analysis
• Complete prediction history view with formatting

📁 Project Structure

Modelisation fin/
│
├── app_mod.py              # Main Streamlit application
├── requirements.txt        # Python dependencies
├── SVR.joblib             # Pre-trained SVR model
├── scaler.joblib          # Scaler for normalization
└── README.md              # This file

🔧 Dependencies

• streamlit: Web interface framework
• joblib: Loading serialized models
• scikit-learn: Machine learning library
• numpy: Numerical computations
• pandas: Data manipulation
• plotly: Interactive visualizations
• matplotlib: Additional plotting

🤖 About the Model

The application uses a Support Vector Regression (SVR) model to perform financial predictions based on:

• Economic indicators (exports, imports, reserves)
• Consumption indices
• Temporal variables
• Price indicators

The model has been pre-trained and serialized to enable fast and efficient predictions.

🎯 Use Cases

• Forecasting financial variables based on economic indicators
• Economic trend analysis
• Exploring relationships between different financial parameters
• Financial decision-making support

📝 Notes

• Predictions are stored in the session history
• History is reset when the application restarts
• Models must be present in the same directory as the application

⚠️ Disclaimer

This system is designed for educational and demonstration purposes. Predictions should not be used as the sole basis for important financial decisions.

📧 Support

For any questions or issues, please create an issue in the project repository.

---

Developed with ❤️ using Python, Streamlit, and Machine Learning