Using News to Predict Stock Movements

This project contains my two submissions for Kaggle competition : Two Sigma: Using News to Predict Stock Movements.
Achieved top 2% in leaderboard.

• Refer to "Featured Notebooks/Analysis/Deliverables" section for fully executed Jupyter Notebooks

-- Project Status: [Completed]

Project Intro/Objective

The purpose of this project is predict a signed confidence value that's correlated with stock price movement.
Therefore, predicted signed confidence value can be used by the competition host to make better decisions on stock trading.

Partner

• Two Sigma is hosting this competition through Kaggle.
• In this competition, market data is provided by Intrinio and news data is provided by Thomson Reuters.

Methods Used

• Exploratory Data Analysis
• Deep Learning
• Data Visualization
• Predictive Modeling

Technologies

• Python
• Jupyter Notebook
• Pandas
• Numpy
• Matplotlib
• Seaborn
• etc.

Project Description

You can find base code in .ipynb format and its executed output by clickin "- executed" link. Data used in this executions are only accessible through Kaggle's kernel, which is a virtual environment within server for copyrights reasons. Therefore, it is not possible to replicate this work.
To learn more about the data, you can view it here.

Featured Notebooks/Analysis/Deliverables

• NN Model.ipynb
• NN Model.ipynb - executed
• XGBoost.ipynb
• XGBoost.ipynb - executed