In short
Machine learning is not magic and not just a buzzword: it is a computer learning patterns from examples and using them to make predictions. Precious Kolawole, a data scientist and machine learning engineer at Lendsqr, makes the idea concrete with a toddler who learns that "red and bright" means pain, and with a classifier trained on labeled cat images. In the Q&A she adds a practical route for beginners: Python first, then data analysis, data science and finally machine learning.
Key takeaways
- Explain machine learning to a non-technical person with a simple everyday analogy.
- Distinguish a model that learns patterns from data from a program built on explicit instructions.
- Describe how a classifier learns a function Y = f(X) from labeled examples during training.
- Recognise the kinds of data-heavy decisions where machine learning adds the most value.
- Plan a beginner learning path from Python through data analysis and data science to machine learning.
Starting from a simple definition
Kolawole opens with a personal challenge. After learning Python, cleaning data and building models, she kept asking herself whether she could explain her work to a complete lay person. Her answer was to strip the subject down to its essentials and explain it "like I'm explaining to a kid."
She begins with a plain definition that separates machine learning from the wider field of AI: "Machine learning is a large subset of artificial intelligence that allows computers to learn from data." (0:32) In her framing, it observes patterns, learns from those patterns and then uses what it has learned to make predictions. The rest of the talk turns that sentence into pictures.
The child, the candle and the stovetop
Her central analogy is a young child seeing a candle flame for the first time. Curious, the child touches it, feels the heat and cries, and the lesson gets stored: something red and bright is painful. Two days later the same child spots a stovetop in the kitchen, notices the same red, bright glow, remembers the candle and keeps away.
That, Kolawole says, is the whole idea in miniature: "To be clear, this is just machine learning because this young child learned pattern from the candle flame to make predictions on the stovetop." (2:19)
She then changes one detail to show what machine learning is not: "Let's now say this young child, he ignored the stovetop because his parents warned him against it. Then that would be explicit programming and not machine learning." (2:19) In traditional programming, a developer writes the instructions and the computer follows them. Machine learning, she explains, flips this: you give the computer inputs and outputs, such as images and their labels, and it builds the model on its own.
What a model is for
To explain what a model does, Kolawole sorts everyday decisions into three groups. Some are easy, like brushing your teeth or paying rent. Some are harder, like asking out a crush or planning a productive day. And some are almost impossible without data, such as which stock to invest in or whether a borrower will repay a loan or default. Machine learning, she argues, shines in that third group.
She sums up its purpose in one line: "And the ultimate goal of machine learning is to simplify large amount of data into a single decision or a prediction." (3:56) The model sits in the middle as a representation of how the real world works on your data, and it is the final product of the whole process.
Teaching a computer to spot cats
The second example makes the mechanics visible. Imagine a program that returns true if an image contains a cat and false if it does not. Instead of writing rules for whiskers and ears, you feed the computer, say, 1,000 cat images labeled true and 1,000 non-cat images labeled false.
Kolawole frames this as an equation, Y = f(X): X is the image, Y is its label, and f is the function the model works out by itself. Learning from the labeled data is training. Once training is done, a new image can be passed through the learned function and the model returns its prediction. "And this entire process is learning." (5:18)
Where beginners should start
The Q&A turns to newcomers asking how to begin. Kolawole is clear that you cannot jump straight into machine learning: "So before you can get started in machine learning, you have to learn Python." (8:09) From there, she describes a common progression into data analysis and visualization (with tools such as Power BI, Seaborn, Tableau and Matplotlib), then data science, and finally machine learning topics such as model building, deployment and deep learning.
Asked about a no-code route, she points to ready-made platforms she believes are available on Microsoft Azure, where you can work with models through a configured interface. Her bottom line is still honest: "But most times for you to dive into machine learning, you know, you must know how to write codes." (10:17)
Watch the talk
"Explain Machine Learning as to a kid" by Precious Kolawole (about 11 minutes): https://www.womentech.net/video/explain-machine-learning-kid-0
About the speaker
Precious Kolawole is a Data Scientist/ Machine Learning Engineer at Lendsqr, a financial company in Nigeria, and is skilled in data preprocessing and analysis, data visualization, supervised machine learning, deep learning, computer vision and web development. She learned to code through online and offline courses, StackOverflow and trial and error, and her interests lie at the intersection of artificial intelligence and healthcare.
Questions this talk answers
- What is machine learning in simple terms?
- How is machine learning different from explicit programming?
- What kinds of decisions is machine learning good for?
- How should a complete beginner get started in machine learning?
- Can you do machine learning without writing code?
Drafted with AI from the talk transcript. Editor review pending.