This project is a team project for the Data Stream Processing's course at Ecole Polytechnique. The goal is to predict the stock market using stream-learning neural networks helped with nlp integration of twitter sentiment estimation.
- BERSANI--VERONI Thomas [email protected]
- GENSBITTEL Luc [email protected]
- LERMITE Titouan [email protected]
To clone the repository, run the following command:
git clone https://github.com/TitouanLMT/M2DS_DataStream/tree/main
The requirements are in the requirements.txt file. You can install them with the following command:
pip install -r requirements.txt
This part assumes you are running the commands in seperate terminals.
To run the project, you need to run a Zookeeper server. To do so, run the following command, assuming you are in the correct Zookeeper directory:
zookeeper-server-start.sh config/zookeeper.properties
To run the project, you need to run a Kafka server. To do so, run the following command, assuming you are in the correct Kafka directory:
kafka-server-start.sh config/server.properties
This script is used to train the model. Run it with the following command:
python model_training_continual.py
This script is used to handle the dataset. Run it with the following command:
python dataset_handler.py
This script is used to make predictions. Run it with the following command:
python rolling_prediction.py
This script is used to mock the stock API. Run it with the following command:
python stock_mock_API.py
The resuls can be seen in the presentation slides of the project.
- [1] Twitter et la bourse: Résumé des articles cités: une corrélation possible, une causalité improbable
- [2] Yamina Tadjeddin, 2013, « La finance comportementale, une critique cognitive du paradigme classique de la finance » Idées économiques et sociales 2013/4 (N° 174), pages 16 et suivantes
- [3] https://www.cairn.info/revue-idees-economiques-et-sociales-2013-4-page-16.htm
- [4] Malcolm Baker & Jeffrey Wurgler, 2007. "Investor Sentiment in the Stock Market," NBER Working Papers, National Bureau of Economic Research, Inc.
- [5] Jean-Christophe Feraudet, 2020. « Analyse de Twitter en temps réel avec Kafka, Spark et mongoDB », https://cedric.cnam.fr/vertigo/Cours/RCP216/docs/UASB03_Projet_Feraudet_v1.0.pdf , le CNAM
- [6] Singh, T., Kalra, R., Mishra, S. et al. An efficient real-time stock prediction exploiting incremental learning and deep learning. Evolving Systems (2022), https://doi.org/10.1007/s12530-022-09481-x