[STUDY GUIDE] • HIGH-YIELD PRACTICE & REVIEW EDITION
CS 7643 Quiz 3 Questions & Answer Key
Solutions Latest 2026/2027 Edition
Comprehensive Examination Question Bank • Concept Mapping
TOTAL QUESTIONS EXAM TOPICS ANSWERS
39 Questions 10 Modules 100% Verified
DOCUMENT OVERVIEW
This document contains 39 questions with correct answers and explanations related to deep learning concepts, specifically
focusing on neural networks and convolutional architectures. It is suitable for students preparing for courses in computer vision,
machine learning, or curriculum mapping, offering a comprehensive review of critical topics necessary for understanding
advanced neural network techniques.
EXAM BLUEPRINT & TOPIC DISTRIBUTION
Systematic breakdown of subject domains and exam coverage.
Topic Module Scope & Core Focus Questions Share (%)
This topic covers various types of errors that can occur when
Neural Network Errors working with neural networks. 3 Qs 7.7%
This topic explores the architecture and functioning of different
Convolutional Neural Networks convolutional neural networks. 5 Qs 12.8%
This topic delves into techniques for visualizing and interpreting
Activation Mapping Techniques neural network activations. 3 Qs 7.7%
This topic examines various loss functions used in training neural
Loss Functions networks. 6 Qs 15.4%
This topic discusses popular object detection models and their
Object Detection Models architectures. 6 Qs 15.4%
Forward and Backward This topic focuses on the concepts of forward and backward
Propagation propagation in convolutional layers. 2 Qs 5.1%
Network Initialization and Transfer This topic covers strategies for initializing networks and the
Learning effectiveness of transfer learning. 2 Qs 5.1%
Confidential • Student Study Edition • Practice & Review Guide Page 1 of 12
, STUDENT STUDY & MASTERY EDITION PRACTICE & REVIEW GUIDE
This topic addresses the concepts of receptive fields and the
Receptive Fields and Parameters calculation of parameters in CNNs. 6 Qs 15.4%
This topic includes methods for understanding and interpreting
Model Interpretability model predictions. 4 Qs 10.3%
Style and Content Loss in Neural This topic explores the concepts of style transfer and content loss in
Networks the context of neural networks. 2 Qs 5.1%
Total Exam Coverage 10 Integrated Topic Modules 39 Qs 100.0%
TOPIC 1: NEURAL NETWORK ERRORS
3 Questions • 7.7% of Exam • This topic covers various types of errors that can occur when working with neural networks.
QUESTION 1
Modeling Error
Answer: Given a particular NN architecture, the actual model that represents the real world may not be in that space.
When model complexity increases, modeling error reduces, but optimization error increases.
QUESTION 2
Estimation Error
Answer: Even if finding the best hypothesis, weights, and parameters that minimize training error, may not generalize to
test set.
QUESTION 3
Optimization Error
Answer: Even if your NN can perfectly model the world, your algo may not find good weights that model the function.
When model complexity increases, modeling error reduces, but optimization error increases.
TOPIC 2: CONVOLUTIONAL NEURAL NETWORKS
5 Questions • 12.8% of Exam • This topic explores the architecture and functioning of different convolutional neural networks.
QUESTION 4
AlexNet
Confidential • Student Study Edition • Practice & Review Guide Page 2 of 12
CS 7643 Quiz 3 Questions & Answer Key
Solutions Latest 2026/2027 Edition
Comprehensive Examination Question Bank • Concept Mapping
TOTAL QUESTIONS EXAM TOPICS ANSWERS
39 Questions 10 Modules 100% Verified
DOCUMENT OVERVIEW
This document contains 39 questions with correct answers and explanations related to deep learning concepts, specifically
focusing on neural networks and convolutional architectures. It is suitable for students preparing for courses in computer vision,
machine learning, or curriculum mapping, offering a comprehensive review of critical topics necessary for understanding
advanced neural network techniques.
EXAM BLUEPRINT & TOPIC DISTRIBUTION
Systematic breakdown of subject domains and exam coverage.
Topic Module Scope & Core Focus Questions Share (%)
This topic covers various types of errors that can occur when
Neural Network Errors working with neural networks. 3 Qs 7.7%
This topic explores the architecture and functioning of different
Convolutional Neural Networks convolutional neural networks. 5 Qs 12.8%
This topic delves into techniques for visualizing and interpreting
Activation Mapping Techniques neural network activations. 3 Qs 7.7%
This topic examines various loss functions used in training neural
Loss Functions networks. 6 Qs 15.4%
This topic discusses popular object detection models and their
Object Detection Models architectures. 6 Qs 15.4%
Forward and Backward This topic focuses on the concepts of forward and backward
Propagation propagation in convolutional layers. 2 Qs 5.1%
Network Initialization and Transfer This topic covers strategies for initializing networks and the
Learning effectiveness of transfer learning. 2 Qs 5.1%
Confidential • Student Study Edition • Practice & Review Guide Page 1 of 12
, STUDENT STUDY & MASTERY EDITION PRACTICE & REVIEW GUIDE
This topic addresses the concepts of receptive fields and the
Receptive Fields and Parameters calculation of parameters in CNNs. 6 Qs 15.4%
This topic includes methods for understanding and interpreting
Model Interpretability model predictions. 4 Qs 10.3%
Style and Content Loss in Neural This topic explores the concepts of style transfer and content loss in
Networks the context of neural networks. 2 Qs 5.1%
Total Exam Coverage 10 Integrated Topic Modules 39 Qs 100.0%
TOPIC 1: NEURAL NETWORK ERRORS
3 Questions • 7.7% of Exam • This topic covers various types of errors that can occur when working with neural networks.
QUESTION 1
Modeling Error
Answer: Given a particular NN architecture, the actual model that represents the real world may not be in that space.
When model complexity increases, modeling error reduces, but optimization error increases.
QUESTION 2
Estimation Error
Answer: Even if finding the best hypothesis, weights, and parameters that minimize training error, may not generalize to
test set.
QUESTION 3
Optimization Error
Answer: Even if your NN can perfectly model the world, your algo may not find good weights that model the function.
When model complexity increases, modeling error reduces, but optimization error increases.
TOPIC 2: CONVOLUTIONAL NEURAL NETWORKS
5 Questions • 12.8% of Exam • This topic explores the architecture and functioning of different convolutional neural networks.
QUESTION 4
AlexNet
Confidential • Student Study Edition • Practice & Review Guide Page 2 of 12
