Using Data Science to Develop Business Intelligence

Master of Science in Information Technology (MSIT) Professor Ed Malthouse talks about the importance of data mining and why knowledge learned in his Northwestern class can be used to inform business decisions.

Buzzwords run rampant in the business world, and today, many of those words have to do with data. 

Big data. 

Data analytics.

Machine learning.

Ed MalthouseData is around us now more than ever before, and the key for businesses is to figure out how to use accessible data to their advantage. In Northwestern Engineering's Data Science for Business Intelligence course, Professor Ed Malthouse teaches Master of Science in Information Technology (MSIT) students techniques for data mining and its use in various business applications to enable business intelligence. The course allows students to gain hands-on experience using state-of-the-art data mining tools to model business problems and discover patterns to use as support for business decisions. 

By the end of the class, students will understand the complete process of using data to make better business decisions, from extraction and cleaning to modeling and presenting. Just as importantly, students will learn the limitations of data.

Malthouse recently took a few minutes to talk about the course and why this knowledge can be a game-changer for students. 

How do you like to describe the course to someone not familiar with data science?

This is a broad survey class that attempts to introduce both supervised and unsupervised learning. Within supervised learning, there are two main classes of problems: one where we attempt to make causal inferences about how inputs affect the output, and the second is focused on associating the inputs as accurately as possible with the output. 

Students will get experience with both types of applications and get exposure to both classical and modern machine learning methods. We also spend a few weeks on unsupervised learning problems of clustering and dimensionality reduction. I also introduce the field of recommender systems. My course gives students an overview of the entire field and gives them a solid foundation to build on.

Your course description says that "in the rapidly changing business environment, with global competition and maturing markets, competitive advantage is extremely important." Why do you feel that's the case?

Having a competitive advantage is important to stay in business. Without it, other firms will eat your lunch. Good use of data can help a firm know the effects of its actions. They can know which actions are working and which ones are not. This leads to a more efficient allocation of resources. They can also use data and computational methods to optimize business processes.

What types of hands-on experiences do students get in your class?

This is an engineering class, so there are weekly problem sets. I think it is important to give problems that come from real business problems with real data. That brings the material to life. At the same time, I think it is important to work simpler problems by hand to understand how the different algorithms work. Doing so gives the student a deeper understanding of the method, its limitations, and what makes it work.

What are some of the data mining tools that students are able to use?

I teach the course in R and Python. R is easier to learn but Python has more job opportunities. Students can pick which one they want.

What are the most important lessons you hope they walk away from your class having learned?

I hope they become passionate about using data and models to inform their business decisions. I also want them to see through the hype and understand that it is difficult to maintain usable data sources, and many things can go wrong when building models. This should make them a more skeptical business person when it comes to making machine learning and modeling investments.

What are one or two of your favorite examples of how businesses have been able to use data mining to their competitive advantage?

One obvious company is Amazon. I’m amazed at how much they have optimized the user experience on their site, and how good their recommendation systems are. I also admire their test-and-learn mentality. Every time you visit Amazon, you’re probably part of a randomized controlled experiment of some sort. 

I’m also very impressed with the recommender systems at Spotify. They help me find new music that interests me, which is of value to me.

What do you enjoy most about the course?

I like the MSIT students very much. They are hard-working and experienced. I also love the material in the class. I’ve been doing research with these models for over 27 years, and there’s a lot of new methods being introduced all the time.

McCormick News Article