Data Mining 2.1

Front

What best captures the goal of machine learning?

Back

Building systems that automatically learn and improve from experience

Answer explanation

Machine learning focuses on creating systems that can learn from data and improve their performance over time without being explicitly reprogrammed. Instead of hard-coding rules, ML algorithms identify patterns and make decisions based on experience gained from training data.

Front

What makes ML different from traditional programming?

Back

It writes its own rules

Answer explanation

1. Traditional Programming

In the traditional way of building software, a programmer:

• Writes explicit rules that the computer follows step-by-step.

• The logic and decision-making are manually coded by humans.

Example:
Imagine you want to make an email spam filter.
In traditional programming, you might write rules like:

“If the subject contains the word ‘free’ and ‘money’, mark it as spam.”

This works for a few cases but quickly becomes unmanageable — spammers can change words, use emojis, or find new tricks.
So every time the pattern changes, you must rewrite the rules.

---

2. Machine Learning Approach

Machine learning flips this process.
Instead of telling the computer how to make decisions, we give it:

• Data (examples) — like thousands of spam and non-spam emails

• An algorithm — a way to learn from those examples

Then, the computer figures out its own rules from the patterns it sees in the data.

Simplified formula:

Traditional programming: Rules + Data → Answers

Machine learning: Data + Answers → Rules

Front

Which of the following tasks would least likely involve machine learning? Grouping similar customers based on purchase history, Filtering unwanted emails from an inbox, Sorting documents alphabetically by filename, Suggesting words while composing a text message

Back

Sorting documents alphabetically by filename

Answer explanation

Machine learning is used for pattern recognition and prediction tasks (like grouping customers, filtering spam, or suggesting words).
However, sorting alphabetically follows a simple rule-based algorithm, not a learning process — so it doesn’t require ML.

Front

Who first popularized the term “Machine Learning”?

Back

Arthur Samuel

Answer explanation

The concept of Artificial Intelligence traces back to Alan Turing in the 1950s, when he asked, “Can machines think?” and proposed the Turing Test to measure it.
But it was Arthur Samuel, an American computer scientist at IBM, who first coined the term “Machine Learning” in 1959.

Samuel created a program that played checkers (draughts) and got better over time by analyzing past games — one of the earliest examples of a computer improving through data rather than explicit programming. He described this as giving a computer the ability to “learn without being explicitly programmed.”

His work laid the foundation for what we now know as supervised learning and self-improving systems, bridging the gap between theory and practice.

Later, in the 1980s and 1990s, figures like Geoffrey Hinton advanced neural networks, and by the 2010s, deep learning became the driving force behind modern AI systems such as image recognition and speech translation.

Front

How are Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) related?

Back

ML is a subset of AI, and DL is a subset of ML

Answer explanation

Artificial Intelligence (AI) is the broad field focused on creating systems that can perform tasks that normally require human intelligence — such as reasoning, problem-solving, perception, and learning.

Within AI, Machine Learning (ML) refers to the set of techniques that allow systems to learn patterns from data rather than being explicitly programmed.

Deep Learning (DL) is a subset of ML that uses multi-layered neural networks to automatically extract complex features from raw data — making it especially powerful for tasks like image recognition, speech processing, and natural language understanding.