WorksheetsGeneral Questions
Total questions: 15
Worksheet time: 8mins
Who published the first theoretical model of an artificial neural network in 1943?
Alan Turing
Warren McCulloch & Walter Pitts
John Searle
Geoffrey Hinton
What question did Alan Turing famously ask in the 1950s?
"Can robots dream?"
"What is intelligence?"
"Can machines think?"
"How to teach a machine?"
What is “Weak AI”?
AI that learns slowly
AI that appears smart but doesn't understand
AI without data
AI that is smaller than Deep Learning
Which of the following is NOT a type of data used in AI?
Numerical
Textual
Image/Video
Political Speech
What happens when a model is "overfitted"?
It performs better on new data
It underestimates test results
It memorizes training data and fails on new data
It ignores training examples
Which type of ML involves learning from labeled input–output pairs?
Supervised
Unsupervised
Reinforcement
Semi-supervised
What is the process of adjusting weights based on error called?
Forward propagation
Overfitting
Backpropagation
Reinforcement
Which is not a source of AI bias?
Data bias
Algorithmic bias
Societal bias
Temperature bias
What makes a neural network "deep"?
Wide input layers
Many hidden layers
Big datasets
GPU usage
Which neural network is best at handling sequential data like text or speech?
Convolutional Neural Network (CNNs)
Feedforward Neural Network
Recurrent Neural Networks (RNNs)
Expert System
What do Convolutional Neural Networks (CNNs) specialize in?
Translating text
Detecting bias
Generating speech
Recognizing spatial patterns in images
What are the 3 main steps in training an AI model?
Collection → Sharing → Selling
Storage → Visualization → Summarizing
Acquisition → Labeling → Training
Download → Testing → Coding
What is "feature extraction" in ML?
Removing useless code
Compressing large files
Encrypting training data
Turning raw data into useful characteristics
Which ML type uses rewards to learn?
Supervised
Unsupervised
Reinforcement
Semi-supervised
What does the “loss function” do in neural networks?
Stores weights
Deletes poor models
Tracks CPU usage
Measures error
