Machine Learning for Dummies Trivia Quiz

Answer: Enabling computers to learn and improve without being explicitly programmed

Machine Learning (ML) is a subset of Artificial Intelligence (AI) that focuses on systems that can learn from experience and data. Instead of following hard-coded instructions, these systems find patterns, make predictions, and improve over time as they are exposed to more information.

For example, an ML model trained on thousands of images of cats can learn to recognize cats in new photos, not because it was told exactly how a cat looks, but because it learned the pattern from data.

Answer: Training Data

Think of it like teaching a student with examples before giving them a test. The student studies past cases, like how house prices change based on size or location, so they can spot patterns.
In the same way, a machine learning model learns from a bunch of past data. This collection of examples is called a training dataset. It's what the model uses to understand how things usually work, so it can make smart guesses when it sees something new.

Answer: Testing Data

The data used after training to measure how well the model performs on completely new, unseen examples is called the testing dataset.

It's like giving a student an exam after they've studied. It checks how well they've actually learned, not how well they memorized practice questions.

Using test data helps us understand if the model can generalize, that is, make good predictions on real-world data rather than just memorizing patterns from training.

Answer: Overfitting

Overfitting happens when a machine learning model learns the training data too well, including noise, random details, and irrelevant patterns, instead of the real underlying relationships. As a result, it performs amazingly on the training set but poorly on new, unseen data.

It's like a student who memorizes past exam answers word-for-word instead of understanding the concepts. They are great at practice tests but terrible at new questions.

Answer: Loss Function

A loss function is like a scoreboard for a machine learning model. Every time the model makes a prediction, the loss function checks how far off it was from the correct answer. The bigger the mistake, the higher the loss.

The model uses this score to learn. It tries to adjust itself (its internal settings or "weights") to reduce the loss next time.

The goal of the model is to keep adjusting internal parameters until the loss is as small as possible, meaning the model is making better predictions.

Answer: Supervised Learning

In supervised learning, the model learns by example. It's given data where each input already has the correct answer, like a photo of a fruit labeled "apple" or "banana." The model studies these examples to figure out patterns, such as what makes an apple look different from a mango.

Once trained, it can look at a new, unlabeled image and make a smart guess, like saying "this looks like a banana". It's called "supervised" because the model learns with guidance, using known answers to build its understanding.

Answer: Unsupervised Learning

Unsupervised learning is when a machine learning model is given data without any answers or labels. It doesn't know what each item is. It just looks for patterns on its own.

For example, it might notice that some customers buy similar things and group them, or that some news articles talk about similar topics and cluster them. The model isn't told what the groups mean. It just finds structure in the chaos.

Answer: Reinforcement Learning

Reinforcement Learning is a type of machine learning where the model learns by doing, like trial and error. It interacts with an environment, makes decisions, and gets feedback in the form of rewards or penalties.

The main purpose is to learn a strategy that earns the most rewards over time. It's used in areas like game-playing, robotics, and self-driving cars, anywhere a system needs to learn from experience and improve its actions.

Answer: Regression

Regression is used when the output is a number that can vary, like price, temperature, or weight. The model learns from past data to find patterns between input features and the target value.
Once trained, it can predict continuous outcomes for new data. For example, it might estimate house prices based on size and location or the weight of a person based on age and height.

Answer: Classification

Classification is used when the output is a category or label, like spam versus non-spam, or different types of disease. The model learns from labeled data to recognize patterns and assign new inputs to the correct group. For example, it might identify whether an email is spam based on its content or the sender.

The goal is to sort data into predefined classes, not predict continuous numbers.