Supervised Learning and Financial Literacy for Teenagers

Are Teenagers Spending Money Wisely?

Are teenagers spending money wisely?

It’s a deceptively simple question. When we ask whether spending is “wise,” we are really asking whether it is an investment or just consumption.

Investment can mean the obvious: putting money into a savings account or purchasing equity shares that might appreciate over time. But it can also mean investing in growth. A tutoring session, music lessons, even the decision to pay for a sports trainer might look like consumption at first glance. Very expensive ones. But under the right lens, they are closer to investments. Consumption, on the other hand, often refers to the purchase of non-durable goods: snacks, clothing, phone accessories, or entertainment.

The distinction is subtle, and sometimes blurry. Even for adults, it isn’t always clear which category a decision falls into. If a teenager buys a pair of running shoes, is that consumption? Or, if those shoes become the appropriate foundation for healthy exercise habits, could it be an investment in long-term well-being?

Supervised Learning in Plain English

This kind of labeling problem, and in our case, whether a decision falls into investment or consumption, is exactly the type of challenge machine learning was built to handle. In particular, it can be addressed using supervised learning, where the goal is to use data to train a model that outputs the right label given the right features. In other words, supervised learning algorithms can help us understand the impact of our financial decisions.

Think about it like this:

• Features: the measurable details of a decision. For example, how much money was spent, what category of item it was (gym fees, video game, tutoring session, slime). Some features are categorical features (e.g., durable vs non-durable, investment vs consumption), others are numerical features (purchase amount, number of repeat purchases).

• Label: the category we ultimately care about. Is it good for our long-term growth? Precisely, in this case, whether the decision should be understood as investment or consumption.

By showing a supervised learning algorithm many examples of spending decisions, paired with the correct labels, the model can learn to recognize patterns. Over time, it improves at predicting the label for new, unseen examples.

But here is the catch: the label itself is not always obvious. For a student athlete, hiring a personal training coach might look like consumption to the untrained eye. In the short run (2–5 years), however, it can lead to improved performance or better school placements. In the long run (>10 years), it can establish habits that compound into better health outcomes or even professional opportunities. What counts as “investment” depends on the time horizon and the definition of return we choose. In short, there are many things to consider!

This illustrates one of the most important lessons in supervised learning: the quality of your model depends not just on the algorithm but on the clarity of your labels. The training signal is only as reliable as the assumptions baked into it.

Why This Matters

When we introduce supervised learning into the conversation about teenager financial literacy, we are doing more than implementing algorithms blindly. We encourage students to critically think about the categories they use in everyday life, to question how they draw the line between short-term consumption and long-term investment. While supervised learning can help us find patterns between consumption choices and future observable outcomes, we also need to be mindful of its limitations and the context of its interpretations.

Reference:

Lecture slide of CPSC 340 Machine Learning and Data Mining by Professors Prajeet Bajpai and Mathias Lécuyer.

Do you want to read more on Machine Learning in Plain English? Comment below or reach out to us to let us know what you think! Over the coming weeks, we might use examples like this to unpack key ideas in machine learning and show how they connect to decisions we make every day.