AI interview questions for freshers, with simple answers

Fifteen basic AI interview questions

1. What is the difference between AI, machine learning and deep learning?
2. What are supervised, unsupervised and reinforcement learning?
3. What is the difference between classification and regression?
4. What is overfitting, and how do you prevent it?
5. Why split data into training, validation and test sets?
6. What is a confusion matrix?
7. Explain precision, recall and F1 score.
8. What is gradient descent?
9. How does linear regression work? Logistic regression?
10. How does a decision tree decide where to split?
11. What is k-means clustering?
12. What is a neural network, and what does an activation function do?
13. What is cross-validation?
14. What is a large language model?
15. Tell me about an AI or ML project you built.

Simple answers to the core questions

• AI vs ML vs DL: AI is the broad goal of machines doing tasks that need intelligence. Machine learning is a way to achieve it by learning patterns from data. Deep learning is machine learning with many-layered neural networks.
• Types of learning: supervised learns from labelled examples (spam or not spam); unsupervised finds structure in unlabelled data (customer segments); reinforcement learns by trial and reward (game-playing agents).
• Classification vs regression: classification predicts a category; regression predicts a number such as a price.
• Overfitting: the model memorises training data, including noise, and fails on new data. Prevent it with more data, simpler models, regularisation, dropout, early stopping and cross-validation.
• Data splits: train to fit, validation to tune choices, test once at the end for an honest estimate.

Evaluation metrics

A classic follow-up: a model is 99% accurate at detecting a disease that affects 1% of people. Is it good? Not necessarily — predicting “healthy” for everyone also scores 99%. That's why precision and recall matter on imbalanced data.

Algorithms in one line each

• Linear regression: fits a straight line (or plane) minimising squared error.
• Logistic regression: outputs a probability for a class using the sigmoid function.
• Decision tree: splits data on the feature that best separates classes, measured by Gini impurity or information gain.
• Random forest: many trees on random subsets, averaged, reducing overfitting.
• k-means: groups points into k clusters by repeatedly assigning to the nearest centre and moving centres.
• k-nearest neighbours: predicts from the k most similar training examples.

Neural networks and LLMs, simply

A neural network is layers of simple units; each computes a weighted sum of its inputs and passes it through an activation function such as ReLU, which adds the non-linearity that lets networks learn complex patterns. Training adjusts the weights with gradient descent and backpropagation to reduce error. A large language model is a very large neural network — a transformer — trained to predict the next token in text, which turns out to produce useful abilities such as answering questions and writing code. It can also produce confident mistakes, called hallucinations.

Talking about your project

Expect most of the interview to be about your project. Explain the problem, the data and where it came from, the model you chose and why, how you evaluated it and the result, and what you would improve. Be honest about what you built versus what came from a tutorial — interviewers ask detailed follow-ups, and “I followed a tutorial for the model, then changed the preprocessing because…” is a perfectly good answer.

Practising

Answer these out loud, in two minutes or less each, and expect a follow-up on every claim. A free AI interview assistant such as CoPilot Interview can run AI fundamentals questions as a mock interview built from your résumé and score each answer. For the wider campus process, see campus placement interviews; for deeper topics, deep learning interview questions.

Disclosure: CoPilot Interview is our product, so weigh what we say about it accordingly. Everything else on this page applies whichever tool you use, or none.

Frequently asked questions

What AI questions are asked to freshers?

AI versus ML versus deep learning, types of learning, classification versus regression, overfitting, data splits, evaluation metrics, common algorithms, neural network basics, LLMs and your own project.

How do I explain overfitting simply?

Overfitting is when a model memorises its training data, including noise, so it performs well on that data but poorly on new data — like a student who memorises answers instead of understanding the subject.

Do freshers need to know deep learning for AI interviews?

Basics, yes: what a neural network is, what activation functions do and how training works. Deep architecture details are usually expected only for specialist roles.

How should I prepare for my first AI interview?

Revise fundamentals with a simple example for each, practise explaining metrics on imbalanced data, and prepare to discuss one project in depth, including what you'd improve.