Course Objectives

This is a practical hands-on course structured as a “teach and do,” where attendees will be presented the technical concepts of machine learning which they will then apply to their data. The objective of the course is for the attendee to learn how machine learning methods can support the following use cases:

• Learn & validate data driven algorithms
• Validate existing rule-based algorithms
• Measure the influence and importance of underlying threat data
• Infer missing or unknown data
• Establish optimal assessment intervals
• Support the assessment of un-piggable pipelines
• Support monetized risk-based decision-making

Who should attend

• Integrity managers & engineers
• Risk managers & engineers
• Data analysts

CEUs

On completion of the course, participants will be awarded 1.4 CEUs.

Preparation to attend

• Install machine learning desktop software on PC (a download link and installation instructions will be provided prior to the course)
• Prepare example data set (template will be provided)
• Watch software instructional videos (<2 hrs)

COURSE NOTES

• Presentation slides and related documentation will be printed in full color in spiral-bound book form as well as PDF download
• Machine learning configuration file (example machine learning processes)

Lecturer

Michael Gloven, P.E. is Managing Partner of Pipeline-Risk (www.pipeline-risk.com) a Denver-based engineering consultancy focused on machine learning and risk management practices for the natural gas, hazardous liquids and water industries. He started his career with Conoco Pipeline holding various operational, engineering and management positions throughout the US. He then went on to co-found Bass-Trigon, a software and engineering company, and as president helped build the organization as a leading provider of risk and cathodic protection data management software. After acquisition of Bass-Trigon, he became an advisor to several international energy and technology companies, and soon afterwards started a new company focused on API 1173 process management for the pipeline industry which was then acquired in 2015. For the last several years, Mike has merged his software, engineering and domain expertise with machine learning practices to advance problem solving in the pipeline industry.