Detecting Invoice Noncompliance Issues through Machine Learning
Day and Time: Info
Field of Study: Basic
Most organizations are using basic criteria when sampling invoices for their compliance reviews and overlooking hidden, more complex patterns of noncompliant behavior in their datasets. This leads to minimal, noncompliant findings and many missed opportunities for recoverables. Kanza Mazhar, E-Commerce Compliance Analyst with Marathon Oil, will talk to us on utilizing machine learning techniques to detect potential overcharges in non-pricebook invoices. She will walk us through challenges with their current compliance review process and lack of meaningful findings that led to the proposal of developing an algorithm that better detects noncompliant behavior in invoices. Kanza will walk us through the steps to developing an accurate algorithm including, the challenges of gathering and preparing invoice data, selecting and improving the final model, and using various combinations of factors to predict what invoices are more likely to be noncompliant.
What you will learn:
- Understand the five-step machine learning process
- Learn about different machine learning algorithms and model improvement techniques
Recognize the key predictors of overcharges specific to Marathon Oil
E-Commerce Compliance Analyst - Marathon Oil
The Oil and Gas Vendor Audit Roundtable is registered with the National Association of State Boards of Accountancy (NASBA) as a sponsor of continuing professional education on the National Registry of CPE Sponsors. State boards of accountancy have final authority on the acceptance of individual courses for CPE credit. Complaints regarding registered sponsors may be submitted to the National Registry of CPE Sponsors through its website: .