Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.


Feature Engineering for Click Fraud Detection

FDMA 2012: International Workshop on Fraud Detection in Mobile Advertising

Clifton Phua

on 18 August 2013

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Feature Engineering for Click Fraud Detection

Background Feature Engineering
for Click Fraud Detection Clifton, Eng Yeow, Ghim Eng, Kelvin, Minh
starrystarrynight Team
Institute for Infocomm Research (I2R) Problem Features Overall Temporal Spatial Potential Methods Results Conclusion Acknow-
ledgments Contact Experience Challenge Meaningful Standing on the
shoulders of giants Rewarding Fraud Publisher Clicks Competition
conditions Data How
many? What? Click behavior Click duplication High-risk
click behavior One minute
duplication Six-hour
periods Countries Publisher's
top-k most
click feature Discretization Generalized Boosted
Models (GBM) Bernoulli
distribution 5000
decision trees 0.001
learning rate depth of 5 5 minimum
observations starrystarrynight is
1st on test set?!? What is the underlying
click fraud scheme? What sort of concealment
strategies commonly used
by fraudulent parties? How to interpret data for
patterns of dishonest
publishers and websites? How to build effective
fraud prevention/detection
plans? SMU BuzzCity Worthy
competitors I2R Clifton Phua
cwcphua@i2r.a-star.edu.sg Questions? Our datasets and details are available at
Full transcript