Deep Learning Unit 1 – Introduction to Neural Networks: important questions for AKTU
Unit 1 (Introduction to Neural Networks) questions that AKTU repeats most often. This unit carries about 30 marks per paper. Start with the repeated questions, then the most asked topics.
Most repeated Unit 1 questions
- Explain the backpropagation algorithm in detail. Derive the weight update rule for a multilayer perceptron. (7 marks, 2026, Training a Network (Loss, Backprop, SGD))
- Derive the mathematical expression for the output of a single hidden-layer neural network. Explain the role of activation functions. (7 marks, 2026, Neural Networks & Shallow Networks)
- Discuss the significance of activation functions in the Universal Approximation Theorem. (7 marks, 2026, Universal Function Approximation)
Most important Unit 1 topic
- Training a Network (Loss, Backprop, SGD) (Unit 1: Introduction to Neural Networks) – asked 10 times in 2023, 2025, 2026
Most asked Unit 1 topics
- Training a Network (Loss, Backprop, SGD) – asked 10 times
- Linear Models (SVM, Perceptron, Logistic Regression) – asked 8 times
- Neural Networks & Shallow Networks – asked 5 times
More Unit 1 previous year questions
- Define model in machine learning. (2 marks, 2026, Introduction to Machine Learning)
- What is the difference between linear regression and logistic regression? (2 marks, 2026, Linear Models (SVM, Perceptron, Logistic Regression))
- Discuss the differences between a perceptron and a support vector machine in terms of learning. (2 marks, 2025, Linear Models (SVM, Perceptron, Logistic Regression))
- What is a loss function, and why is it essential for training a neural network? (2 marks, 2025, Training a Network (Loss, Backprop, SGD))
- Discuss the differences between a perceptron and a support vector machine in terms of decision boundary formation. (2 marks, 2025, Linear Models (SVM, Perceptron, Logistic Regression))
Unit 1 syllabus topics
- Introduction to Machine Learning
- Linear Models (SVM, Perceptron, Logistic Regression)
- Neural Networks & Shallow Networks
- Training a Network (Loss, Backprop, SGD)
- Universal Function Approximation