How Humans Learn?

Humans learn in three main ways:

Learning from an expert
Using knowledge from experts in new situations
Learning by ourselves through trial and error

Let’s look at each type with simple examples.

1. Learning from an Expert

This is when someone with experience teaches us directly.

A baby learns to say "hand" or "blue" because their parents told them.
In school, children learn letters, numbers, and more complex subjects like math and science from teachers.
In college or training, students learn specialized skills from instructors — for example, doctors study anatomy and engineers study physics.
At work, new professionals are guided by experienced colleagues or mentors.

In all these cases, learning happens because someone with knowledge passes it on.

2. Using Expert Knowledge in New Ways

Sometimes we use what we've learned from experts in different situations.

A child who was taught colors can group objects by color, even if no one told them to do that.
A student can tell the difference between nouns and verbs because their teacher explained it before.
At work, someone might choose which customers to target using advice they got earlier from a boss.

This kind of learning isn’t direct. It uses past lessons to make decisions in new situations.

3. Learning by Yourself

Sometimes we have to figure things out on our own.

A baby learns to walk around obstacles after falling many times.
A child learns to ride a bike or an adult learns to drive a car — often through trial and error.
We also learn from mistakes and build a mental list of what works and what doesn’t.

Self-learning is about gaining knowledge through personal experience.

Summary

We learn in three main ways:

From teachers or experts,
By applying what we’ve learned from experts in new situations, and
Through self-experience and trial and error.

Each type plays an important role in how we grow and develop skills throughout life.

How Do Machines Learn?

Machines learn in three basic steps:

Data Input – Using past data to make future decisions
Abstraction – Finding patterns or rules from the data
Generalization – Applying those patterns to new, unseen situations

Let’s understand this by comparing it with how humans learn.

Comparing Machine Learning with Human Learning

Imagine students preparing for exams. At first, they may try to memorize everything. This works when the subject is small and questions are simple. But as topics grow more complex and questions become harder, memorizing isn't enough.

What do good learners do? They focus on key ideas, build a conceptual map, and use that to answer any question — even ones they haven’t seen before.

For example, instead of memorizing every animal and its traits, students can group animals into categories like mammals, birds, reptiles, etc., and remember the key features of each group.

This makes learning easier and more effective.

Abstraction in Machine Learning

In machine learning, data is the starting point. But machines don’t just store raw data — they create a model, which is like a concept map.

A model could be:

A set of if/else rules
A mathematical equation
A tree or graph structure
A way to group similar things together

Choosing the right type of model depends on:

The problem being solved (e.g., prediction, classification)
The nature of the input data (quality, type, completeness)
The domain (e.g., banking, healthcare)

Once a model is chosen, it needs to be trained using data. For example, if the model is an equation like y = c + mx, we need to find the values of c and m based on real data. This process is called training, and the data used is called training data.

Generalization in Machine Learning

After training, the next step is generalization – using the model to make decisions about new data.

But here’s the challenge:

The model was trained on limited data.
New data might have characteristics the model hasn’t seen before.

So, just like humans sometimes rely on intuition when facing something unknown, machines also use approximate methods to handle new cases.

This comes with risks — both machines and humans can make mistakes when relying on guesses rather than facts. But generalization allows both to deal with the unknown in smart ways.

Summary

Machines learn through:

Input – Using data to start the learning process
Abstraction – Creating models that capture patterns from the data
Generalization – Applying those models to make decisions on new data

Like humans, machines move from memorizing facts to understanding concepts and applying them creatively. This is what makes machine learning powerful and flexible.