Ajay Kumar Garg Engineering College, Ghaziabad
DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING
ARTIFICIAL INTELLIGENCE(KCS071)
[1] What is reinforcement learning?
A type of ML where agents learn by taking actions to maximize rewards in an
environment.
[2] Explain the concept of overfitting.
When a model performs well on training data but poorly on unseen data due to
excessive complexity.
[3] What is underfitting?
When a model is too simple to capture the underlying patterns in the data.
[4] What is a confusion matrix?
A table used to evaluate classification models, showing true positives, false positives,
true negatives, and false negatives.
[5] Define precision and recall.
Precision: True Positives / (True Positives + False Positives)
Recall: True Positives / (True Positives + False Negatives)
[6] What is a perceptron?
A basic unit of a neural network that performs binary classification based on input
weights.
[7] What is gradient descent?
An optimization algorithm that adjusts weights in a model to minimize the loss
function.
[8] What is the role of activation functions? List some common activation
functions?
They introduce non-linearity to the model, enabling it to learn complex patterns.
Examples: Sigmoid, ReLU, Tanh, Softmax.
[9] Name some popular AI frameworks.
TensorFlow, PyTorch, Keras, Scikit-learn, and OpenAI Gym.
[10] What is TensorFlow and PyTorch?
An open-source ML library for numerical computation and building AI models. A
deep learning framework known for its flexibility and dynamic computation graphs.
[11] What is a pre-trained model?
A model trained on a large dataset that can be fine-tuned for specific tasks.
[12] What are common challenges in AI?
Data quality, ethical considerations, computational power, and explainability.
[13] What is explainable AI (XAI)?
AI systems designed to provide human-understandable explanations for decisions.
[14] How do you handle imbalanced datasets?
Techniques include oversampling, undersampling, or using weighted loss functions.
[15] What is data augmentation?
Generating additional training data by applying transformations to existing data.
[16] How do you evaluate an AI model?
Using metrics like accuracy, precision, recall, F1-score, or mean squared error.
[17] What are ethical concerns in AI?
Bias in algorithms, privacy issues, job displacement, and accountability for decisions.
[18] What is cross-validation?
A technique to evaluate model performance by splitting the dataset into training and
testing sets multiple times.
[19] What is a convolutional neural network (CNN)?
A neural network designed for image processing tasks, utilizing convolutional layers.
[20] What is a recurrent neural network (RNN)?
A type of neural network for sequential data, such as time series or text.
