The Curiosity of a Child: A Surprising Foundation for Data Science

If you asked me to explain data science to a 5-year-old, I’d probably say something like this: “It’s like being a detective! You collect clues (data), figure out how they fit together (analyze and clean the data), and use them to solve mysteries or guess what might happen next (predict outcomes).”

Data science isn’t just for detectives or grown-ups with fancy degrees. Chances are, you’ve probably applied data science principles in your everyday life even without realizing it. Take me, for instance.

Growing up, I was a curious, quiet child who never stopped thinking. On the outside, I didn’t say much, but in my head, it was a whole market. I connected dots nobody else seemed to notice and asked endless questions, and not just the typical “Why is this like this?” but the extra inquisitive ones: “When will it happen? What might make it happen?” Even now, as an adult, I have a bad habit of asking too many questions. A friend once told me, “You ask questions like a child!” I took it as a compliment.

It was this curiosity that led me to unknowingly embrace data science principles as a child, particularly when it came to something as unpredictable as NEPA (now PHCN). If you’ve ever lived in Nigeria, you know that power outages were so common that light coming on felt like a mini celebration. I hated the unpredictability of it all so, I started paying attention. I took it upon myself to crack the NEPA code.

My notebook became my mini database. I observed the patterns of NEPA’s behaviour, and recorded every time the power went off and when it came back on. (Data collection). I’d come home from school and immediately ask my mom or whoever was home, “Did they bring the light? When did they take it?” I’d write down what I noticed - how long the light stayed, what time it came on, and when it went off. It was like a ritual; I needed to know the "status update" on the power supply.

Over time, I factored in patterns and external factors like public holidays, festive seasons, weather conditions, football matches., etc It took some time, but I started noticing trends. For example, if there was a heavy downpour, I just knew they wouldn’t bring the light for hours because, well, the wires needed time to “dry.” Or if Nigeria was playing a football match, chances were high that NEPA would feel generous. If I were building the NEPA model today, I’d have to account for more complex variables- like the number of times the national grid collapses (which feels like every other week).

Using these patterns, I created a "mental model.” As unpredictable as the power supply was, I could still predict it to an extent. My siblings and neighbours started treating me like I had insider information and I earned the unofficial title "NEPA forecaster". They’d ask me, “When will the light come back?” and I’d confidently respond, “Give them two hours; it’ll come on.”

Sure, the data wasn’t perfect. NEPA "fell my hands" a few times. Sometimes, the power didn’t come back when I thought it would, or it went off unexpectedly. But for the most part, my predictions were surprisingly accurate. So, I filtered out these inconsistent cases (Data cleaning) and focused on finding patterns.

Looking back, I realize that I was applying core data science processes all along:
Data Collection: I gathered information about the power supply—when it came, when it went, and how long it stayed.
Data Cleaning and Preparation: I removed irrelevant details and focused on key variables, like weather conditions or the time of day.
Exploratory Data Analysis (EDA): I looked for patterns in my notes to understand how the power supply worked.
Data Modeling: I created a “mental model” to predict when the power would come back based on the patterns I’d identified.
Model Evaluation: I tested my predictions against reality. If the light didn’t come on when I expected, I adjusted my model.
Model Deployment: My “model” became useful to others—my siblings and neighbours relied on my forecasts to make decisions.

As funny as this sounds, this childhood experiment with NEPA was my first taste of data science. Data science is more than just numbers, charts, and algorithms. It’s about solving real-world problems, just like I did with my power predictions as a child. Yes, data science can be complicated, but at its core, it’s just structured curiosity - something we all have within us. And for me, it all started with NEPA and a notebook.

Over time, we’ll explore not just the different concepts and processes of data science, but also the problems we can solve and the solutions we can create. Whether it’s analyzing a dataset or finding insights to improve lives, data science is a tool we can all use to make better decisions.

Let’s dive in together - one clue, one pattern, and one prediction at a time!

The above is the detailed content of The Curiosity of a Child: A Surprising Foundation for Data Science. For more information, please follow other related articles on the PHP Chinese website!