Stock Market Prediction using Machine Learning

Introduction

We built a ML model to predict France’s stock market, trained on France’s CAC40 dataset from Yahoo Finance! We implemented SVM for Portfolio Optimization for Trend Prediction, and developed LSTM-models fusing datasets to improve prediction & made a 7-day prediction.

Objectives

1. Firstly, a study on the various applications of Machine Learning in finance was provided. This would help to create a more general picture of how Machine learning and finance are connected.
2. Secondly, the work includes a study of the portfolio management problem as well as applying the SVM and neural network methods in the French stock market.
3. Thirdly, a study on the credit risk evaluation problem as well as ”overdraft” data. Also an applica.tion that takes into account all types of standard customer data.

Overview of the Repository

In this repo, you'll find :

• LSTM: Code for LSTM Vanilla & advanced models in jupyter notebook
• Data: Dataset of France stock market 2020 from Yahoo Finance
• data_processing : Code for data pre-processing and feature engineering for SVM model for financial indicators
• models : Code for SVM model weekly & monthly results, to invest or not
• Reports : Detailed documentation of our approach, software and results

Getting Started

1. Clone our repo: git clone https://github.com/HusseinLezzaik/Stock-Market-Prediction.git
2. Install dependencies:

   conda create -n stock-market python=3.8
   conda activate stock-market
   pip install -r requirements.txt
   

3. Run LSTM_METHOD.ipynb for the method you want to make stock market prediction

Collaborators

Hussein Lezzaik, Denis Demko, Thomas Deroo, Doris Fejza, Elona Karaj, Estia Maliqari, Yijue Xie.

Acknowledgments

Our work was built on top of Haifei Zhang master thesis work at UTC-France, read more here.