The Rotman Review: Preparing For The Future Of Machine Learning by: Dr. Peter Zhang on August 22, 2021 | 981 Views August 22, 2021 Copy Link Share on Facebook Share on Twitter Email Share on LinkedIn Share on WhatsApp Share on Reddit The third industrial revolution began to take form near the end of the 20th century. Here, skills that supported digitization or automation became highly valued. With the demise of household products such as analog devices, many companies and professionals risked being left behind. To get ahead, I believed an MBA would equip me with skills relevant to artificial intelligence (AI) and machine learning. Machine learning is a subset of artificial intelligence that involves building models with datasets. Depending on the algorithms that power such models, users are able to leverage machine learning for predicting outcomes or for making decisions. One classical example in business is when a bank decides whether it should give out a loan. This is a process that can be entirely automated by a machine learning algorithm. With algorithms being completely data-driven, lending decisions can be more objective than those made by a human bank manager. Case in point: a study published in 2018 found that algorithms on average charge minority borrowers 40% less than face-to-face lenders. Dr. Peter Zhang FALLING INTO PYTHON’S GRIP AI and machine learning have already disrupted a number of industries. For example, social media giants like Twitter use machine learning to personalize your feed. Virtual assistants like Alexa have harnessed AI technology to cater to the individual needs of its users. And popular rideshare services like Uber uses machine learning to minimize the wait time of its passengers. Although its current usage is impressive, the future of AI is even more exciting. From predicting patient outcomes to developing self-driving cars, it is clear that as future leaders we need be prepared for the innovations to come. As an MBA student, when it was time to pick my electives, I thought the best way to gain AI exposure was to take “Machine Learning In Finance” taught by Professor John Hull. I had greatly anticipated diving into this complex topic. Despite having very little background in computer programming, I was keen to learn Python. Today, Python has grown to become one of the most popular programming languages, with 8.2 million users across the world. For Python users, there is already a large collection of packages intended to support the building and testing of machine learning models. Furthermore, beginners often find Python easier to dive into compared to languages like C or Java. This combination of features makes Python a common choice among data scientists. As a beginner myself, I was grateful for the support from my peers and teaching staff in the “Machine Learning In Finance” class. Equipped with a dedicated Python help desk, the elective made every effort to make learning accessible. MANY APPLICATIONS TO BUSINESS Throughout the elective, I was exposed to all the ways machine learning can solve problems in finance using programs written in Python. In one class, we talked about unsupervised learning which focuses on recognizing patterns from datasets to build groups of data points known as clusters. The exercise was designed to categorize a list of countries into low, medium, or high-risk clusters based on their features. Features are variables that reflects a characteristic of the data and helps determine the relative distance between data points. We used features like GDP and index scores on peace, corruption, and legal risk to describe individual countries. A visual example of the distance between data points is a graph where countries with similar GDP values are plotted closer to each other. The k-means clustering algorithm was used to allocate countries into the aforementioned clusters. The k-means clustering algorithm creates the three clusters by trying to find a position for its cluster center where the distance between each data point and its cluster center is minimized. Conceptually, the algorithm is trying to draw imaginary lines in the most optimal way to separate the clusters from each other. In a business setting like marketing, one could see how this could be a powerful tool when conducting market segmentation where customers are divided into different subgroups for strategic targeting purposes. The relevance of machine learning in business is endless. From real estate valuations to options trading, Rotman MBA students worked on a variety of projects that put machine learning applicability in the context of the real-world setting. Rotman student at work with Toronto in the background INTERDISCIPLINARY COLLABORATIONS Inspired by the exposure from this elective, I continued to build my Python proficiency through courses offered by FinHub, Rotman’s center for financial innovation including machine learning and AI. But finance is not the only industry affected by machine learning. Just last year, the new Temerty Centre for AI Research and Education in Medicine at the University of Toronto was launched to double down on the commitment to machine learning research. In this respect, one aspect that makes the University of Toronto unique is its diverse range of multidisciplinary faculties where icollaboration is enabled and encouraged. I was keen to partake in the interdisciplinary projects made possible by the University environment as an opportunity to explore the applicability of my machine learning knowledge. With the support of funding from the University of Toronto, I joined a team of MBA, pharmacy, medicine, and law students to launch a machine learning model to predict preterm labour, a major cause of newborn mortality. Our project, “Roo”, intends to improve childbirth outcomes, especially in areas with insufficient healthcare resourcing. Not only is this an opportunity to support patients in need, it also allows me to develop skills in project management for machine learning innovations as an MBA student. Beyond the university environment, the machine learning investments in the surrounding institutions in Toronto fosters innovation in AI technology. This passion for AI innovation has sparked not only academic fervor, but has also attracted significant commercial interest. As the number of AI startups continues to expand, so do the opportunities to immerse yourself in the business of launching new disruptive technologies. TORONTO IS THE PLACE TO BE Trang Nguyen is an MBA Candidate at Rotman School of Management and a student Fellow at the Creative Destruction Lab. This summer she is working with ODAIA, a Toronto-based startup focused on solutions driven by machine learning. At Rotman, the Creative Destruction Lab (CDL), is an incubator for startups with no equity requirements. MBA students who pursue a fellowship have the opportunity to partner with experienced founders to work together to commercialize the technology. CDL also offers its MBA fellows specialized courses that teach them how to grow startups. Examples of startups that have graduated from CDL’s AI stream include BenchSci, Deep Genomics, and Darwin. By harnessing both the classroom and practical opportunities offered by the CDL, student fellows become skilled in finding the right product market fit for emerging companies with powerful ideas. Trang Nguyen is a MBA candidate of the class of 2022 and one such CDL fellow. This summer, she is interning with a Toronto-based startup, ODAIA, which leverages machine learning to drive customer engagement. It currently has two products: Maptual, which offers personalized customer engagements for pharmaceutical firms, and Multitud, a tool for generating insights for direct-to-consumer and e-commerce brands. “I chose Rotman precisely because of CDL, and I was not disappointed,” Nguyen explains. “Thanks to my CDL course, I could understand the technical aspects and the AI algorithms behind my company’s platform, and then use my storytelling and marketing skills to translate that “monster” into bite-sized, simple, and digestible business language.” Toronto as a city has an important role in all the intensity surrounding AI innovation. Not only does it have the highest concentration of AI startups, the University of Toronto’s academic prowess attracts and fosters AI talent from all over the world. In fact, the machine learning algorithm that powers ODAIA had been developed by Dr. Perikilis Andritsos and Dr. Gael Benard, who are researchers at the University of Toronto. Canada as a nation has also been highly supportive of AI growth. It was the first country to launch a national strategy on AI and has also produced the highest number of AI patents per capita when compared to China and other G7 countries. As AI and machine learning move to the forefront of technological innovation, it is important for us to learn the language of data science and the skills necessary to work with data scientists. Similar to the disruptive nature of the digital revolution, many risk being left behind. In order to stay competitive, now is the time to invest in a deeper dive in machine learning, and for this, there is no place I would rather be than in Toronto. Bio: Dr. Peter Zhang, PharmD is an MBA candidate at the University of Toronto’s Rotman School of Management located in the heart of Toronto. Through the unique combined Doctor of Pharmacy/MBA degree program, he has explored the intersection between life sciences and commercial strategy. Additionally, he has published research works in peer-reviewed academic journals and opinions in national media outlets in Canada.