1 AIT Asian Institute of Technology

Talkingdata adtracking fraud detection challenge

AuthorKumar,Gudapuri Nikhil
Call NumberAIT RSPR no.ICT-21-01
Subject(s)Machine Learning
Fraud--Detection
Data mining--Computer programs
NoteA research submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Information and Communication Technologies, School of Engineering and Technology
PublisherAsian Institute of Technology
AbstractFraudulent clicks is one of the major issues that internet advertise companies are facing. The inefficiency due to this fraud clicks are one of the critical issue that effects advertiser’s revenue. This results to affect the other players who rely on these publishers who pay less to advertise on various platforms. TalkingData manages to receive around 3 billion advertisement clicks every day and 90% of them are potentially fraudulent clicks. This is the issue which we are trying to solve by filtering and distancing out fraud clicks vs reliable clicks every second. The goal of this project is to create a Machine Learning algorithm that identifies and distinguishes fraud clicks by creating a patterns/Fingerprint using various attributes such as IP, app id, device id, OS Id, and channel id. After analyzing the data, we can distinguish unusual clicks recurring from the same fingerprint without downloading mobile applications, it will be flagged as a fraudulent click. We will feature the provided dataset and apply various Machine Learning algorithm and compare them.
Year2021
TypeResearch Study Project Report (RSPR)
SchoolSchool of Engineering and Technology (SET)
DepartmentDepartment of Information and Communications Technologies (DICT)
Academic Program/FoSInformation and Communication Technology (ICT)
Chairperson(s)Teerapat Sanguankotchakorn
Examination Committee(s)Bohez, Erik L. J.;Nicole, Olivier
Scholarship Donor(s)Asian Institute of Technology Fellowship
DegreeResearch Studies Project Report (M. Eng.) - Asian Institute of Technology, 2021


Usage Metrics
View Detail0
Read PDF0
Download PDF0