Academic Highlights 

The second half of the MMA program began in March and things picked up quickly. We thought that the final month of the intro term back in January was intense, but looking at all the deadlines for the end of April to the beginning of May, I almost miss January.

Luckily, I have wonderful teammates, and working through problems together has been a rewarding process – it is always interesting to see how many ways there are to tackle the same problem.

Now, onto the course specifics. The second half of the program focuses on applications and if I had to pick a theme, I would say it highlights how simple problems can be difficult to solve in practice. A great example is the supply chain class with Professor Andre Cire. When we talk about well-known topics in class such as the traveling salesman problem (TSP), the solutions seem straightforward. But once you begin optimizing your model with inventory, capacity, timing, and many other constraints, this problem quickly becomes incredibly complex. Soon, you have generated what feels like hundreds of subtours and somehow, your model remains “infeasible”.

The upside is that you are never alone in the program. If you are stuck after multiple attempts, other students are probably experiencing something similar.

Our class has arranged tutorials and even created a Discord channel with the professor to provide additional support.

Other highlights include building natural language processing models and image recognition models for the AI course with Professor Brian Keng. The course provides an excellent intro to relevant and interesting deep learning topics. Another example is working on detecting earnings management, fraud, and predicting returns in the accounting course with Professor Scott Liao. I have worked with time-series data before, but this course has provided more context for working with financial data which is useful because finance is a big industry in Toronto.

Reflections on the program

With the wealth of online resources, you can self-learn most topics. I decided to join the MMA program because I wanted to learn best practices, be able to ask questions, and not have to worry about the accuracy of the information being taught. 

As we have learned time and time again, it is easy to perform poorly thought-out analysis and jump to the wrong conclusions. The projects I mentioned above are good examples of what I enjoy most about the program. I like the process of going over the statistical/theoretical basis behind different methods then solidifying this knowledge by applying concepts and justifying the thought process. Why would you pick one model over another? What are the benefits, limitations, and risks associated with this analysis?

Now that I have been through most of the program, I feel confident in my ability to make well-informed decisions at all stages of the analytics pipeline from design to ETL (Extract, Transform, Load) to model testing. There is no way the program can teach you everything, the world of data science is too big and time is a big limitation.

 I feel equipped to dive into technical topics on my own and my list of interesting data science topics/tools to explore post MMA grows daily!

The Master of Management Analytics is designed to give students the advanced data management, analytics, and communication skills needed to become an analytics professional.