/LimitOrderBook

Stock market predictions based on order books.

Primary LanguageJupyter NotebookMIT LicenseMIT

LimitOrderBook

Stock market predictions using limit order books dataset. Based on this paper.

Table of contents

General info

Given LOB (Limit Order Book) data and certain points in time our goal is to predict whether the price will increase or decrease.
We will be using LOB data from London stock market, collected for September 2013.
Main method used is Logistic regression.

Screenshots

Example screenshot

Technologies

  • Python - version 3.7.3

Libraries

  • numpy
  • pandas
  • sklearn
  • scipy
  • csv
  • matplotlib
  • seaborn

Results

  • 51.6 % baseline on test data
  • around 54% accuracy on train data
  • around 55% accuracy on test data

Status

Project is: finished,

Inspiration

Data Mining class.

Authors

Created by @TheFebrin and @MatMarkiewicz