https://github.com/zrkhadija/multivariate-sector-price-prediction-using-macroeconomic-indicators

This project focuses on Multivariate Sector Price Prediction using macroeconomic indicators. By leveraging a custom-built LSTM model in PyTorch, we predicted the prices of 10 financial sectors simultaneously. The model takes as input the historical price data of these sectors along with key macroeconomic indicators.
https://github.com/zrkhadija/multivariate-sector-price-prediction-using-macroeconomic-indicators

data-preprocessing deep-learning hyperparameter-tuning multivariate-timeseries optuna pytorch pytorch-implementation time-series-analysis timeseries-forecasting timeseriesforecasting

Last synced: 13 days ago
JSON representation

This project focuses on Multivariate Sector Price Prediction using macroeconomic indicators. By leveraging a custom-built LSTM model in PyTorch, we predicted the prices of 10 financial sectors simultaneously. The model takes as input the historical price data of these sectors along with key macroeconomic indicators.

• Host: GitHub
• URL: https://github.com/zrkhadija/multivariate-sector-price-prediction-using-macroeconomic-indicators
• Owner: zrkhadija
• Created: 2024-11-21T15:07:38.000Z (about 1 month ago)
• Default Branch: main
• Last Pushed: 2024-11-21T15:16:01.000Z (about 1 month ago)
• Last Synced: 2024-12-02T16:25:51.305Z (20 days ago)
• Topics: data-preprocessing, deep-learning, hyperparameter-tuning, multivariate-timeseries, optuna, pytorch, pytorch-implementation, time-series-analysis, timeseries-forecasting, timeseriesforecasting
• Language: Jupyter Notebook
• Homepage:
• Size: 9 MB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# 📈 Multivariate Sector Price Prediction Using Macroeconomic Indicators

This project leverages deep learning to predict the prices of 10 financial sectors simultaneously. By combining historical sector data and macroeconomic indicators, a custom-built LSTM model in PyTorch was developed from scratch to handle multivariate time series data.

# Project Highlights

- ** Advanced Preprocessing Technique**: insuring stationnarity without losing memory of information from original data.

- **Multivariate Predictions**: Predict the prices of 10 financial sectors simultaneously.

- **Custom LSTM Model**: Built from scratch using PyTorch to handle complex time series relationships.

- **Macroeconomic Indicators**: Integrated key economic data with sectoral historical prices for improved forecasting.

- **End-to-End Workflow**: Includes data preprocessing, model development, and evaluation.

  # 📊 Dataset Description

  The inputs for this project include:

- **Sector Historical Data**: Price data for 10 financial sectors over time.

- **Macroeconomic Indicators**: Features such as GDP, inflation rates, unemployment rates, and other relevant indicators.

  The combined dataset captures both sectoral trends and macroeconomic influences, essential for accurate multivariate predictions.

  # 🛠️ Technologies Used

- **🐍 Python**: For data manipulation and analysis.

- **🔗 PyTorch**: For building the LSTM model.

- **📊 pandas/numpy**: For preprocessing and data handling.

- **📉 Matplotlib/Seaborn**: For visualizing predictions and analysis results.

- **📉 Optuna**: for Hyperparameter tunning.

# ✨ Insights and Applications

- **Financial Forecasting**: Use the model to predict sectoral trends based on historical data and economic factors.

- **Risk Management**: Aid in understanding sectoral dependencies and economic impacts.

- **Model Generalization**: Extend the approach to other markets or industries.