If you’re a data enthusiast, marketing expert, or entrepreneur, you’re probably no stranger to the thrill of diving into complex datasets, uncovering insights, and solving problems creatively. My journey through Block 5 of the Master of Data Science (MDS) program at the University of British Columbia (UBC) was a perfect blend of challenges, insights, and moments of pure “aha!”. Let me take you along as I share what I learned, how these courses enriched my perspective as both a data scientist and a marketing expert, and how these experiences might resonate with your own data-driven pursuits.
DSCI 553: Bayesian Inference and Computation II
One of the most intellectually enriching courses I took during this block was Bayesian Inference and Computation II. In this course, we pushed beyond the basics and immersed ourselves in understanding the underlying mechanics of inferential statistics, bridging theoretical concepts with computation. I found myself particularly excited about how Bayesian Inference came into play here. Given my background in digital marketing and market research, the ability to predict customer behavior using probability distributions felt like magic—but it was magic rooted in data, numbers, and sound statistical principles.
A highlight for me was using Python to conduct Markov Chain Monte Carlo (MCMC) simulations, giving a practical spin to complex concepts. This added a layer of depth to how I now approach data modeling in my professional projects. One thing I’ve learned, especially in the world of digital campaigns, is that data is never just “random.” It’s influenced by underlying systems, and this course gave me the tools to uncover those systems.
DSCI 563: Unsupervised Learning
Another fascinating part of Block 5 was DSCI 563: Unsupervised Learning. We tackled more advanced unsupervised learning models, working with clustering, dimensionality reduction, and recommendation systems. I loved seeing how different models performed in the face of various datasets—a real example of no one-size-fits-all. In the world of entrepreneurship and marketing, personalization is key, and the concept of recommendation systems made me reflect on how digital products could adapt dynamically to user behavior in real-time. Imagine a marketing funnel that evolves based on your unique clicks and views. This is where the beauty of unsupervised learning for customer experience truly comes to light.
My favorite project involved using clustering techniques to group customer data—it brought me back to my days at Lingano, our language learning startup, where I managed content categorization and audience segmentation. Understanding the process behind how models “group” and classify gave me a deeper insight into tailoring experiences for our users.
DSCI 531: Data Visualization II
Let’s not forget about the storytelling side of data—DSCI 531: Data Visualization II. This course was all about effectively communicating data-driven stories to audiences, which is something I’ve always been passionate about. Whether it’s presenting to investors or crafting marketing reports, I’ve come to realize that data visualization is the bridge that connects raw insights with human understanding.
I particularly enjoyed designing interactive dashboards using Tableau and Plotly, transforming static datasets into actionable insights for business stakeholders. During my time at Bookapo, a book-summary platform, we often needed to make quick strategic decisions based on audience engagement data. A visual dashboard is worth a thousand spreadsheets—and this course sharpened my skills in creating visualizations that are not only beautiful but purposeful.
DSCI 574: Spatial and Temporal Models
The ability to work with both spatial and temporal data has always fascinated me, and DSCI 574: Spatial and Temporal Models did not disappoint. We explored techniques such as time series decomposition, ARIMA models, and geospatial analysis, which are invaluable for understanding patterns that evolve over time and space. In marketing, it’s all about understanding where and when to act, and this course gave me new tools to assess which strategies directly impact outcomes over time.
As Head of Growth at InstaHop, my role often revolves around analyzing campaign performance—what works, what doesn’t, and why. Understanding spatial and temporal dynamics is at the core of growth marketing, and the tools I gained through this course have already started to impact how I set up campaigns, interpret results, and iterate on strategies.
Key Takeaways from Block 5
- Bayesian Thinking in Marketing: Bayesian Inference showed me how to apply Bayesian concepts to customer segmentation and predictive modeling.
- Personalized Customer Experience: Unsupervised Learning gave me ideas about using recommendation systems to adapt marketing strategies dynamically.
- Data Storytelling: Effective visualizations are not just for data scientists—they’re for anyone who needs to make informed decisions quickly.
- Spatial and Temporal Analysis: Understanding both spatial and temporal data is crucial for attributing success to specific actions, not just relying on intuition.
- Reproducibility Matters: Establishing workflows that are scalable and reproducible can make or break long-term projects.
Final Thoughts: Applying These Lessons in Real Life
Block 5 at UBC was a powerful combination of technical skills, practical projects, and strategic insights. For those of you who are data enthusiasts, marketing experts, or entrepreneurs, I can’t stress enough how data science is becoming the bedrock for making smarter decisions—whether it’s optimizing a marketing campaign, creating personalized customer experiences, or scaling a business with data-backed insights.
If you’re intrigued by any of the topics I covered—whether it’s the technical underpinnings of Bayesian inference or practical applications in marketing—reach out to me! I’d love to connect, share more insights, and learn about how you’re incorporating data science into your world.