Foundations of Modern Predictive Analytics

Learn data mining techniques and tools that will allow you to make the link between business needs and your technical skills. This intensive Hands-On course is taught by James Drew who has over 30 years experience and developed by the largest organization of analytics professionals, INFORMS.


What You'll Learn

As a course participant, you will learn the concepts and practice of a selection of Modern Predictive Analytics (MPA) techniques that you can immediately apply in your everyday analytic life. Through a blend of technical concepts, business examples and real problems, you will learn how to make the link between business needs and your technical skills.

The best way to make this link between business problems and MPA techniques is to actively perform MPA in real business scenarios. The use of readily available commercial software, which you will download onto your laptop prior to class, will allow you to experience the practical effects of MPA concepts and their business application. Although this is neither a software course nor an extensive MPA techniques overview, you will learn how to apply R (a widely used open-source package) to a span of problems.

You will also have ample problem solving practice to prepare you to use your new software and analytic expertise in your own work.

Specifically, this course will help you to:

  • Incorporate Modern Predictive Analytics into your business practices
  • Recognize appropriate methods from business discussions
  • Be a technical resource and MPA advocate in your organization  

Course Outline

Day 1: Putting the “Modern” in Predictive Analytics for Business

I. The MPA Ecosystem

  • Automated techniques and complex data make MPA Modern
  • A mesh of business questions MPA can answer
  • The process of MPA
  • Debunking hype

II. “What do I get for a 1-pt rise….?”: Classic Numerical Prediction

  • Using basic prediction machinery
  • Scoring operational improvements—the “1-pt rise”
  • Finding explainers in wide datasets
  • Common business uses

III. Prediction in a Complicated World: Finding Sweet Spots

  • Finding “sweet spots” of performance in a bumpy world
  • Regression Trees and how to control them
  • Avoiding the perils of automated analysis
  • Finding better models

IV. Making the Links Between Problems and MPA

  • Practicing with real problems
  • Turning words into analytic solutions

Day 2: Developing an MPA Repertoire and Using It

V. Thresholds Between Good and Bad

  • Predicting Good and Bad, the classical way
  • Using odds in real life
  • Assessing and improving classifications

VI. Sweet Spots, Extended

  • Decision trees: a tool for understanding your data
  • Finding pockets of Good and Bad in complicated datasets
  • Focusing on the good explainers

VII. Finding Patterns

  • Using association rules to find fragments of knowledge
  • Deciding “What goes with what?”
  • Distinguishing interesting rules

VIII. Recognizing When to Use All This

  • Confronting real problems
  • Taking the next steps in MPA mastery
  • Incorporating MPA skills in your job
  • Becoming a resource for your organization  


James Drew, Worcester Polytechnic Institute,  James Drew Verizon (ret.)

James Drew was a Fellow at Verizon Corporation for over 30 years, where he was an internal consultant in statistics and data mining until his recent retirement. He has been won such internal honors as the Chairman's Leadership Award and multiple Excellence Awards. During his career, Jim has worked in the areas of survey analysis, quality control, workforce forecasting, customer relationship management, repair optimization and legal consulting, among others.

Jim has also been an adjunct associate professor in the Management Department of Worcester Polytechnic Institute for over 15 years. He is a Chartered Statistician of the Royal Statistical Society, and appeared on the International List of the 50 Most Prolific Authors in Services Marketing. He is the inventor or co-inventor in 17 patents. In the winter, he is a nordic ski instructor.  



Members - $1,275
Non-members - $1,475
*Discounts available for 3 or more from the same organization.

 Course Dates and Locations

The 2017 schedule is being finalized.  Please contact Bill Griffin, Continuing Education Program Manager, with any inquiries about the 2017 schedule.  


Cancellations - cancellations must be in writing and received 21 days or more prior to the start of the course. A refund will be issued less a $100 processing fee. Cancellations less than 21 days prior to the start of the course will not be eligible for a refund.

Substitutions - if you cannot attend you may send a substitute without incurring a fee provided notice is given in writing (please include substitutes' name) at least 72 hours prior to the start of the course.

Transfers - you may transfer to an earlier or later course date provided the request is received in writing at least 14 days prior to the start date of the course originally booked. There will be a $50 rebooking fee.

Students say...

"As a business analyst and project manager, this course on Modern Predictive Analytics (MPA) not only fulfilled, but exceeded my expectations. Its real-world MPA business and industry examples are something I was looking for. It presents analytics science usefulness for business analysis; not only in decision making, but as a tool for explaining past and present business dynamics and processes, among others. The course has a very good balance between lectures and in-class exercises, with an engaging and experienced instructor who made it all easy to learn. I am looking forward to future INFORMS-sponsored courses like this one."

- Jaime Santiago, PhD, MBA, PMI-PBA, PMP

Course Dates