Authors

Yajun Wang

Type

Text

Type

Dissertation

Advisor

Brusco, Sandro | Zhou, Yiyi | Liu, Ting | Tan, Wei.

Date

2016-12-01

Keywords

Economics | FPDS, public procurement, SVM

Department

Department of Economics

Language

en_US

Source

This work is sponsored by the Stony Brook University Graduate School in compliance with the requirements for completion of degree.

Identifier

http://hdl.handle.net/11401/77411

Publisher

The Graduate School, Stony Brook University: Stony Brook, NY.

Format

application/pdf

Abstract

Fraud in public procurement is a big problem in public sector all over the world, and is very difficult to detect in empirical studies. Common fraud schemes include corruption, collusive bidding, failure to provide the required quality, and false statements, etc. In this paper, I employed game theory, machine learning, and statistical methods to detect fraud risk in Federal Procurement Contract Data, and studied the relationship of fraud, competition and contract types. In the first section, I studied a procurement game and found that if the firms' types are close enough to each other, their strategies regarding whether or not to engage in fraud would tend to be similar. Based on this proposition, in the second section, I implemented One-Class Support Vector Machine method to train the historical data of contractors with fraud records, and developed a classifier. Then I used the classifier to classify and analyze the Federal Procurement Data. In the last section, I applied Logit Regression to the classification outcomes, and the result shows that competition has a small positive relationship with fraud risk. In addition, performance based contracts and flexible-price contracts are more inclined to fraud. | 25 pages

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.