Mexico has kept electronic records of all taxable transactions since 2014 . Anonymized data collected by the Mexican federal government comprises more than80 million contributors (individuals and companies) and almost 7 billion monthly invoices among contributors between January 2015 and December 2018 . With our methods, weestimate previously undetected tax evasion in the order of \$10 billion USD peryear by about 10 thousand contributors . We use twomachine-learning methods to classify other contributors as suspects of taxevasion: deep neural networks and random forests . We thus obtain a list of highly suspicious contributors sorted by the amount of evaded tax, valuable information for the authorities tofurther investigate illegal tax activity in Mexico . We estimate previously untaxed tax evasion is at least $10 billion per year by about ten thousand contributors. We further reduce the number of suspects by focusing on those with a short network distance from known tax evaders. We therefore obtain a number of suspected tax evasions. We use this information to help the authorities in our analysis of tax evasion to identify more than 30,000 people. We hope to use this data to further investigate the authorities’s tax evasion as a tool to identify and prevent further tax evasion by identifying and identifying tax evading individuals and companies that have been identified by the authorities.

Author(s) : Martin Zumaya, Rita Guerrero, Eduardo Islas, Omar Pineda, Carlos Gershenson, Gerardo Iñiguez, Carlos Pineda

Links : PDF - Abstract

Code :

https://github.com/pylablanche/gcForest


Coursera

Keywords : tax - evasion - contributors - authorities - identifying -

Leave a Reply

Your email address will not be published. Required fields are marked *