This repository contains python codes and results of the master thesis "Identification and Modelling of Ultra Fast Extreme Events in Financial Markets: Mini Flash Crashes"

The final document is not published here yet.

• The codes you can find here are:

  • Cleanin_function.ipynb

    Cleaning routine based on Barndorff‐Nielsen, Hansen et al. 2009. It works with data from WRDS TAQ database

  • Drift_detection.ipynb

    Code to identify drift bursts

  • flash_crash_stats.ipynb

    Descriptive statistics and flash crashes identification

• Documents you can find:

  • Flash_crash_Statistical_Learning.pdf

    Essay about outlier identification techniques by MOM procedures in statistical learning literature and application of the Nanex routine

  • Master_Thesis.pdf

    Final document