An Enjoyable Journey with 100+ Forecasters

The next offering of my load forecasting courses is scheduled on March 9-11, 2016 in Charlotte. Registration is open here: https://support.sas.com/edu/schedules.html?id=1326

Among all the professional commitments, I love teaching the most. I teach all over the places, from industry trainings to university lectures, from conference workshops to private onsite classes. This blog post is about an enjoyable journey of teaching my course, "Electric Load Forecasting: Fundamentals and Best Practices" over the last year.

1. An idea

After working on load forecasting projects for several years with many utilities, I realized that I had never got a comprehensive, formal and rigorous classroom training on this subject.
In fact, I was not even able to find such a training from universities nor industry organizations. In 2010, the idea of developing a load forecasting course came to my mind.

At that time, I was with a utility consulting firm. I primarily used SAS and MS Excel for my forecasting projects. Meanwhile, I found that SAS had an education program called Business Knowledge Series (BKS). All of the BKS instructors were renowned industry experts. Most of them were not SAS employees. The BKS courses were more about business knowledge than operating the software. After talking with SAS Education, I started the course development. Although half of the contents were from my PhD dissertation, it still took me about one and a half years to develop the course from an idea.

2. Startup

Since the first class in May of 2012 at SAS World Headquarter, this course has been offered 10 times, to 106 attendees from 42 organizations. Thanks to my co-worker Ms. Rain Xie, who helped me create part of the infograhics below to summarize what has been going on with the course during the last year.

3. Customization

The course was originally designed as a three-day course. For practical reasons, we decided to teach the materials in two days. To fit the three-day materials into a two-day schedule without compromising the quality, I did two things:

1) Limit the class size.
I decided to limit each public offering to no more than 10 people or 6 groups. This may look controversial from the business perspective. The overhead cost is basically fixed. The more people we put in a classroom, the more profit SAS will make. However, due to the diverse background of the attendees, it requires a lot of time to answer the questions from different sectors of the industry for different business needs. Some questions may not be much relevant to other people. In addition, large class size also requires more time helping each attendee go through the hands on exercises. After teaching this course a few times, I found the limit at 10 people or 6 groups for public offerings. For onsite offerings, I did not set any constrains on class size, because all the attendees are from two or three groups of the same company. Instead, I encourage large class size to formulate comprehensive discussions about the business needs.

2) Customize the materials.
I force myself not to give the same speech twice. The only time I strictly followed the course notes was in the first offering. Since then, I've been conducting an attendee profile survey a few weeks prior to each offering. This way I can have an idea about the background of each attendee. Due to the small class size, every offering has a different mix of attendees. Every time I customize the course materials based on the attendees' background. Sometimes, I also add one or two presentations of my recent research projects and findings. For onsite classes, I offer an additional day of fully customized materials. This can be the applications of my methodology to this utility's data, and/or a mini project on several specific topics of interest.

It has been a very enjoyable journey so far.

The next class in Chicago is already full with attendees from three IOUs, two ISO/RTO's and one trading firm. There are three more offerings this fall. I hope to see you in a future class!