����ֱ��

A discussion with Faculty Director Kai Larsen

Looking to pursue a career in business analytics? With constant advancements in technology and analytical tools, the opportunities for harnessing big data continue to grow—as does the value of professionals in this occupation. The Leeds MS in Business Analytics prepares students for a data-driven future rooted in ethical business.

Faculty Director Kai Larsen shared his insights on artificial intelligence in today’s business world and how Leeds is preparing students to harness data and build ethical models that propel businesses forward.  

Engineering decisions from data

Designed to imitate the decision-making ability of a human expert, an expert system uses data to solve particular problems. While a consultant on Norwegian banking early in his career, Professor Larsen gained exposure to expert systems and the power of using data to make decisions.  

“In those early days, we would interview experts like loan processing officers to build ‘if–then’ statements, and focus on developing systems that built that person’s knowledge into technology,” said Larsen. “So essentially, we were addressing the question of: Can you automate the process of giving loans based on, not the data, but the officer?”

The revolution in analytics that has happened since then, explained Larsen, has been in shifting the focus to collecting, integrating, and analyzing the data about the behavior of consumers before and after receiving loans. Artificial intelligence algorithms, properly configured, take care of the rest.

To Larsen, using analytics in this way means the ability to solve problems in novel ways and fuels efficiency through automated processes. The potential for positive impacts, however, is balanced out by a fair amount of negative possibilities—which is why Larsen finds it crucial that analysts are trained to think critically about data.  

Artificial intelligence and consequence ethics

After years of collecting data from algorithms that were created to behave like experts, professionals now have a plethora of evidence they can examine in order to unearth issues and enhance automated systems.

“We have an A.I. model of, for example, who deserves a loan or not. That can be torn apart and examined. Well, it turns out that these algorithms—just like the humans they’re based on—are biased. And that you’ve now automated a biased process.”

Further illustrating the complexity of analytical models, Larsen outlined how a system is comprised of and impacted by both nonexistent and existent data. For example, if loan officers preferred to provide loans to people of a certain race, ethnicity or gender, then the data set and automated system created does not represent the groups of people who were less likely to receive loans. 

“Initially, a person might argue that not having that data could be a good thing. So now, those groups of people who didn’t proportionally receive loans can simply apply and the identifying factor that once prevented them from getting a loan, such as their race, ethnicity or gender, can just be omitted from the data—so the algorithm can’t figure out what that identifying factor is.”

However, a society’s systems and all issues within those systems are deeply connected. External influences and circumstances, all rooted in social identities, help construct people’s individual opportunities and realities. For artificial intelligence, this means that omitting an identifying factor from a data set in order to help right a historical wrong is not that simple.

“Depending on your social identities, you will have had different opportunities in life. Maybe you grew up in a certain area because of your race or ethnicity, which impacts the education you received, what career you have and whether or not you own your house. All of these individual data points, from the products you buy to the car you drive, give hints about your social identities. This makes it very difficult to develop algorithms properly and to prevent any unintentional biases.”

In other words, Larsen explained, the more data you have and the more varied that data, the harder it is to create fair algorithms.

The Leeds approach to business analytics

Training students to not only understand the field of business analytics but also the larger issues in the world is what Larsen finds most exciting about his role. As this discipline continues to evolve, he believes it’s crucial to incorporate ethics when teaching tomorrow’s leaders.

“Today, artificial intelligence is changing the things we can do in a way that’s creating all kinds of challenging and interesting situations. So, training students well enough that they can chart a course through these issues is to me the exciting part. It’s no longer just about training them to be able to create artificial intelligence solutions—it’s about also training them to do so ethically.”

The Leeds MS in Business Analytics is a ten-month program, where students can choose to learn on campus or online. Shaping students into data experts, the program first teaches how to capture, analyze and translate data sets for actionable insights, before diving into artificial intelligence and social issues.

Fortunately, this grey area is where engineering and social sciences are coming together in unprecedented ways. Students in the program can expect to not only become experts in knowing how to deal with data and drive value for businesses, but they can also anticipate being challenged to think critically about the data and its connections to larger world issues.

“We’re not engineers, we’re not computer scientists—that’s not our goal. Our goal is to teach all of these things in the context of business. With every class the students take, it’s about learning the statistics, the machine learning tools and the coding—but always with a goal toward ethical business. It’s also a fairly diverse program, which allows for greater opportunities to talk about social issues.”

Business analysts and their growing value in the workplace

By teaching students the skills necessary to go from data to value, the MS in Business Analytics program at Leeds prepares them to create change in a variety of workplaces—from healthcare to consulting. With this versatile skill set and the growing opportunities for companies to use big data, business analysts have a bright future in the workplace.

“Can you learn how to take all the data that’s available right now in a company, figure out what’s actually important in that data, and create a model in a way that can drive value for a company? That’s how we train our students. But the actual value comes in when a company makes changes based on what you discover from the model.”

To learn more about Faculty Director Kai Larsen’s unique insights and his story, read From Norway to Boulder: A professor’s journey in business analytics.