Unit-2 Advanced concept of Modelling in AI
Revisiting AI, ML, and DL
Revisiting AI, ML, and DL (20 with Answers)
1. Artificial Intelligence (AI) refers to:
a) Machines performing tasks that require human intelligence
b) Only computer programming
c) Human brain research only
d) Internet speed improvement
Answer: a) Machines performing tasks that require human intelligence
2. Which of the following is a subset of AI?
a) Machine Learning
b) Database Management
c) Operating System
d) Cloud Computing
Answer: a) Machine Learning
3. Machine Learning (ML) is best defined as:
a) Making computers learn from data without being explicitly programmed
b) Writing long codes for every task
c) Creating human-like robots only
d) Speeding up internet processing
Answer: a) Making computers learn from data without being explicitly programmed
4. Deep Learning (DL) is a subset of:
a) Machine Learning
b) Database Systems
c) Cloud Computing
d) Operating Systems
Answer: a) Machine Learning
5. Which technique is inspired by the structure of the human brain?
a) Neural Networks in Deep Learning
b) Relational Databases
c) Cloud Servers
d) Excel Spreadsheets
Answer: a) Neural Networks in Deep Learning
6. In AI hierarchy, the correct order is:
a) AI → ML → DL
b) ML → AI → DL
c) DL → ML → AI
d) ML → DL → AI
Answer: a) AI → ML → DL
7. AI is mainly concerned with:
a) Mimicking human intelligence in machines
b) Only storing large amounts of data
c) Faster internet browsing
d) Making games only
Answer: a) Mimicking human intelligence in machines
8. Which of the following is an application of Machine Learning?
a) Spam email detection
b) Writing essays
c) Saving files in folders
d) Turning on WiFi
Answer: a) Spam email detection
9. A key difference between ML and traditional programming is:
a) ML learns patterns from data, while traditional programming follows fixed rules
b) ML requires no data at all
c) ML cannot solve real-world problems
d) ML is slower than traditional programming
Answer: a) ML learns patterns from data, while traditional programming follows fixed rules
10. Deep Learning requires:
a) Large amounts of data and high computational power
b) No data at all
c) Only manual coding for every step
d) Internet access always
Answer: a) Large amounts of data and high computational power
11. Which of the following is a real-life example of Deep Learning?
a) Face recognition in smartphones
b) Saving contacts in phonebook
c) File compression
d) Copy-pasting text
Answer: a) Face recognition in smartphones
12. AI systems that beat humans in chess are examples of:
a) Narrow AI
b) General AI
c) Super AI
d) No AI
Answer: a) Narrow AI
13. Which of these is the MOST advanced level of AI (still theoretical)?
a) Artificial Super Intelligence (ASI)
b) Narrow AI
c) General AI
d) Traditional Programming
Answer: a) Artificial Super Intelligence (ASI)
14. ML algorithms are classified into:
a) Supervised, Unsupervised, and Reinforcement Learning
b) Hardware and Software
c) Databases and Networks
d) Simple and Complex codes
Answer: a) Supervised, Unsupervised, and Reinforcement Learning
15. In Supervised Learning, the model is trained with:
a) Labeled data
b) Unlabeled data
c) No data
d) Only images
Answer: a) Labeled data
16. In Unsupervised Learning, the model works with:
a) Unlabeled data
b) Only labeled data
c) Human-coded rules
d) No data at all
Answer: a) Unlabeled data
17. Reinforcement Learning is based on:
a) Trial-and-error with rewards and punishments
b) Pre-written programs only
c) Database queries
d) Cloud storage
Answer: a) Trial-and-error with rewards and punishments
18. Which of the following AI technology is most useful for self-driving cars?
a) Deep Learning
b) File Management
c) Word Processing
d) Database Sorting
Answer: a) Deep Learning
19. Which of these is NOT an application of AI/ML/DL?
a) Voice assistants (Alexa, Siri)
b) Predicting stock markets
c) Online fraud detection
d) Copying files manually from one folder to another
Answer: d) Copying files manually from one folder to another
20. The relationship between AI, ML, and DL can be shown as:
a) AI is the broadest field, ML is a part of AI, and DL is a part of ML
b) ML is the broadest field, AI is inside ML, and DL is inside AI
c) DL is the broadest field, ML is inside DL, and AI is inside ML
d) AI, ML, and DL are unrelated
Answer: a) AI is the broadest field, ML is a part of AI, and DL is a part of ML
Common Terminologies Used with Data
Common Terminologies Used with Data (20 with Answers)
1. Data can be best defined as:
a) Raw facts and figures without context
b) Meaningful information
c) Processed knowledge
d) Final conclusion
Answer: a) Raw facts and figures without context
2. Processed data is called:
a) Information
b) Knowledge
c) Intelligence
d) Dataset
Answer: a) Information
3. A collection of related data stored together is known as:
a) Dataset
b) Algorithm
c) Code
d) Model
Answer: a) Dataset
4. Each row in a dataset is called a:
a) Record or Instance
b) Feature
c) Column
d) Attribute
Answer: a) Record or Instance
5. Each column in a dataset is called a:
a) Feature/Attribute
b) Record
c) Example
d) Observation
Answer: a) Feature/Attribute
6. Which of the following represents the dependent variable in data analysis?
a) Target/Label
b) Feature
c) Record
d) Dataset
Answer: a) Target/Label
7. Independent variables in a dataset are also known as:
a) Features/Attributes
b) Labels
c) Instances
d) Records
Answer: a) Features/Attributes
8. The process of dividing data into training and testing sets is called:
a) Data Splitting
b) Data Mining
c) Data Cleaning
d) Data Visualization
Answer: a) Data Splitting
9. Which dataset is used to train a machine learning model?
a) Training Set
b) Testing Set
c) Validation Set
d) Output Set
Answer: a) Training Set
10. Which dataset is used to evaluate a trained machine learning model?
a) Testing Set
b) Training Set
c) Input Set
d) Data Pool
Answer: a) Testing Set
11. The unwanted or irrelevant information in data is called:
a) Noise
b) Features
c) Label
d) Record
Answer: a) Noise
12. Converting raw data into a usable format is called:
a) Data Preprocessing
b) Data Visualization
c) Data Storage
d) Data Mining
Answer: a) Data Preprocessing
13. Missing values in data are also known as:
a) Null Values
b) Labels
c) Features
d) Targets
Answer: a) Null Values
14. The process of finding patterns in large datasets is called:
a) Data Mining
b) Data Splitting
c) Data Recording
d) Data Cleaning
Answer: a) Data Mining
15. Which term refers to the range of possible values a variable can take?
a) Domain
b) Record
c) Label
d) Dataset
Answer: a) Domain
16. The step of removing duplicate or incorrect records from data is called:
a) Data Cleaning
b) Data Splitting
c) Data Visualization
d) Data Mining
Answer: a) Data Cleaning
17. Structured data is usually stored in the form of:
a) Tables (rows and columns)
b) Images and videos
c) Free text
d) Audio files
Answer: a) Tables (rows and columns)
18. Unstructured data includes:
a) Text, images, audio, and video
b) Tables and spreadsheets
c) Rows and columns only
d) Database records
Answer: a) Text, images, audio, and video
19. A subset of data selected for analysis or model building is called a:
a) Sample
b) Feature
c) Record
d) Target
Answer: a) Sample
20. The complete collection of all possible data points is known as:
a) Population
b) Sample
c) Dataset
d) Record
Answer: a) Population
Modelling in AI
Modelling in AI (20 with Answers)
1. What does modelling in AI refer to?
a) Building machines
b) Creating mathematical or logical representations of real-world problems
c) Making 3D structures
d) Writing computer hardware codes
Answer: b) Creating mathematical or logical representations of real-world problems
2. The main purpose of modelling in AI is to:
a) Replace humans completely
b) Simplify complex problems for analysis and prediction
c) Create robots only
d) Collect large amounts of data
Answer: b) Simplify complex problems for analysis and prediction
3. Which of the following is an example of modelling?
a) Predicting tomorrow’s weather using past data
b) Sending an email
c) Playing a video
d) Writing a book
Answer: a) Predicting tomorrow’s weather using past data
4. Which step comes first in AI modelling?
a) Data Collection
b) Model Testing
c) Model Deployment
d) Prediction
Answer: a) Data Collection
5. In AI, models are trained using:
a) Datasets
b) Robots
c) Only software
d) Games
Answer: a) Datasets
6. The accuracy of an AI model depends on:
a) The size and quality of data
b) The speed of the computer only
c) The number of robots available
d) Hardware type only
Answer: a) The size and quality of data
7. A model that predicts outcomes based on past examples is called a:
a) Predictive Model
b) Reactive Model
c) Static Model
d) Mechanical Model
Answer: a) Predictive Model
8. In AI, “training a model” means:
a) Feeding the model with data so it can learn patterns
b) Teaching a human
c) Running the final prediction
d) Deploying the program
Answer: a) Feeding the model with data so it can learn patterns
9. Testing a model is done to:
a) Check its performance on unseen data
b) Add more features to it
c) Make hardware stronger
d) Replace datasets
Answer: a) Check its performance on unseen data
10. Which of the following is NOT a type of AI model?
a) Classification Model
b) Regression Model
c) Prediction Model
d) Manual Model
Answer: d) Manual Model
11. A classification model is used when the output is:
a) Categories or classes
b) Continuous numbers
c) Hardware design
d) 3D objects
Answer: a) Categories or classes
12. A regression model is used when the output is:
a) A continuous numerical value
b) A class label only
c) A fixed category
d) None of the above
Answer: a) A continuous numerical value
13. Which model would be best to predict house prices?
a) Regression Model
b) Classification Model
c) Clustering Model
d) Reinforcement Model
Answer: a) Regression Model
14. Which model would be best to identify spam or not-spam emails?
a) Classification Model
b) Regression Model
c) Clustering Model
d) Reinforcement Model
Answer: a) Classification Model
15. Clustering models are used to:
a) Group similar data points without pre-defined labels
b) Classify data into known groups
c) Perform only regression tasks
d) Replace datasets
Answer: a) Group similar data points without pre-defined labels
16. Which of the following is an example of clustering?
a) Grouping customers based on buying patterns
b) Predicting temperature tomorrow
c) Predicting salary from experience
d) Classifying fruits as apple or orange
Answer: a) Grouping customers based on buying patterns
17. In supervised learning models, the data used for training is:
a) Labelled
b) Unlabelled
c) Missing
d) Random
Answer: a) Labelled
18. In unsupervised learning models, the data used for training is:
a) Unlabelled
b) Always labelled
c) Cleaned and labelled
d) Predefined
Answer: a) Unlabelled
19. The real-world deployment of an AI model means:
a) Using the model in actual applications
b) Creating test datasets only
c) Collecting more data
d) Removing the model
Answer: a) Using the model in actual applications
20. The final step of modelling in AI is:
a) Deployment and Monitoring
b) Data Cleaning
c) Feature Selection
d) Data Collection
Answer: a) Deployment and Monitoring
Modelling in AI
Modelling in AI (20 with Answers)
1. What does modelling in AI refer to?
a) Building machines
b) Creating mathematical or logical representations of real-world problems
c) Making 3D structures
d) Writing computer hardware codes
Answer: b) Creating mathematical or logical representations of real-world problems
2. The main purpose of modelling in AI is to:
a) Replace humans completely
b) Simplify complex problems for analysis and prediction
c) Create robots only
d) Collect large amounts of data
Answer: b) Simplify complex problems for analysis and prediction
3. Which of the following is an example of modelling?
a) Predicting tomorrow’s weather using past data
b) Sending an email
c) Playing a video
d) Writing a book
Answer: a) Predicting tomorrow’s weather using past data
4. Which step comes first in AI modelling?
a) Data Collection
b) Model Testing
c) Model Deployment
d) Prediction
Answer: a) Data Collection
5. In AI, models are trained using:
a) Datasets
b) Robots
c) Only software
d) Games
Answer: a) Datasets
6. The accuracy of an AI model depends on:
a) The size and quality of data
b) The speed of the computer only
c) The number of robots available
d) Hardware type only
Answer: a) The size and quality of data
7. A model that predicts outcomes based on past examples is called a:
a) Predictive Model
b) Reactive Model
c) Static Model
d) Mechanical Model
Answer: a) Predictive Model
8. In AI, “training a model” means:
a) Feeding the model with data so it can learn patterns
b) Teaching a human
c) Running the final prediction
d) Deploying the program
Answer: a) Feeding the model with data so it can learn patterns
9. Testing a model is done to:
a) Check its performance on unseen data
b) Add more features to it
c) Make hardware stronger
d) Replace datasets
Answer: a) Check its performance on unseen data
10. Which of the following is NOT a type of AI model?
a) Classification Model
b) Regression Model
c) Prediction Model
d) Manual Model
Answer: d) Manual Model
11. A classification model is used when the output is:
a) Categories or classes
b) Continuous numbers
c) Hardware design
d) 3D objects
Answer: a) Categories or classes
12. A regression model is used when the output is:
a) A continuous numerical value
b) A class label only
c) A fixed category
d) None of the above
Answer: a) A continuous numerical value
13. Which model would be best to predict house prices?
a) Regression Model
b) Classification Model
c) Clustering Model
d) Reinforcement Model
Answer: a) Regression Model
14. Which model would be best to identify spam or not-spam emails?
a) Classification Model
b) Regression Model
c) Clustering Model
d) Reinforcement Model
Answer: a) Classification Model
15. Clustering models are used to:
a) Group similar data points without pre-defined labels
b) Classify data into known groups
c) Perform only regression tasks
d) Replace datasets
Answer: a) Group similar data points without pre-defined labels
16. Which of the following is an example of clustering?
a) Grouping customers based on buying patterns
b) Predicting temperature tomorrow
c) Predicting salary from experience
d) Classifying fruits as apple or orange
Answer: a) Grouping customers based on buying patterns
17. In supervised learning models, the data used for training is:
a) Labelled
b) Unlabelled
c) Missing
d) Random
Answer: a) Labelled
18. In unsupervised learning models, the data used for training is:
a) Unlabelled
b) Always labelled
c) Cleaned and labelled
d) Predefined
Answer: a) Unlabelled
19. The real-world deployment of an AI model means:
a) Using the model in actual applications
b) Creating test datasets only
c) Collecting more data
d) Removing the model
Answer: a) Using the model in actual applications
20. The final step of modelling in AI is:
a) Deployment and Monitoring
b) Data Cleaning
c) Feature Selection
d) Data Collection
Answer: a) Deployment and Monitoring
Types of Machine Learning Models
Types of Machine Learning Models (20 with Answers)
1. How many main types of Machine Learning models are there?
a) Two
b) Three
c) Four
d) Five
Answer: b) Three
2. Which of the following is NOT a type of Machine Learning?
a) Supervised Learning
b) Unsupervised Learning
c) Reinforcement Learning
d) Mechanical Learning
Answer: d) Mechanical Learning
3. Supervised learning uses which type of data?
a) Only images
b) Labelled data
c) Unlabelled data
d) Random data
Answer: b) Labelled data
4. In unsupervised learning, the training data is:
a) Labelled
b) Unlabelled
c) Always numerical
d) Pre-tested
Answer: b) Unlabelled
5. Reinforcement learning is inspired by:
a) Teacher-student learning
b) Trial-and-error learning
c) Copy-paste learning
d) Fixed rule-based learning
Answer: b) Trial-and-error learning
6. Which of the following is an example of supervised learning?
a) Predicting house prices
b) Grouping customers by spending habits
c) Training robots with rewards
d) Clustering flowers by petal size
Answer: a) Predicting house prices
7. Which of the following is an example of unsupervised learning?
a) Identifying spam emails
b) Grouping customers into segments
c) Predicting rainfall amount
d) Predicting salary
Answer: b) Grouping customers into segments
8. Which of the following is an example of reinforcement learning?
a) Teaching a robot to walk using rewards and penalties
b) Predicting exam scores
c) Grouping animals by characteristics
d) Classifying emails as spam or not spam
Answer: a) Teaching a robot to walk using rewards and penalties
9. In supervised learning, the output variable is also known as:
a) Reward
b) Label or target
c) Cluster
d) Feature
Answer: b) Label or target
10. Which type of learning is used in market basket analysis?
a) Supervised Learning
b) Unsupervised Learning
c) Reinforcement Learning
d) All of the above
Answer: b) Unsupervised Learning
11. Which learning type is most suitable for self-driving cars?
a) Supervised Learning
b) Unsupervised Learning
c) Reinforcement Learning
d) None of the above
Answer: c) Reinforcement Learning
12. Classification and Regression problems are solved using:
a) Unsupervised Learning
b) Supervised Learning
c) Reinforcement Learning
d) Random Learning
Answer: b) Supervised Learning
13. Clustering problems are solved using:
a) Supervised Learning
b) Unsupervised Learning
c) Reinforcement Learning
d) Hybrid Learning
Answer: b) Unsupervised Learning
14. Which type of machine learning model learns without any teacher or labelled data?
a) Supervised
b) Unsupervised
c) Reinforcement
d) None
Answer: b) Unsupervised
15. In reinforcement learning, the agent improves its performance through:
a) Only labelled data
b) Rewards and penalties
c) Predefined rules only
d) Random guessing
Answer: b) Rewards and penalties
16. Predicting whether an email is spam or not is an example of:
a) Classification (Supervised Learning)
b) Clustering (Unsupervised Learning)
c) Reinforcement Learning
d) None of these
Answer: a) Classification (Supervised Learning)
17. Predicting tomorrow’s temperature is an example of:
a) Regression (Supervised Learning)
b) Clustering (Unsupervised Learning)
c) Reinforcement Learning
d) Rule-based Learning
Answer: a) Regression (Supervised Learning)
18. Grouping news articles into categories without predefined labels is an example of:
a) Classification
b) Regression
c) Clustering
d) Reinforcement
Answer: c) Clustering
19. Which model type is best when we want to train without any output labels?
a) Supervised Learning
b) Unsupervised Learning
c) Reinforcement Learning
d) None
Answer: b) Unsupervised Learning
20. Which of the following describes supervised learning best?
a) Learning from labelled data with correct answers provided
b) Learning from unlabelled data without guidance
c) Learning by rewards and punishments
d) Learning without using data
Answer: a) Learning from labelled data with correct answers provided
Types of Deep Learning
Types of Deep Learning (20 with Answers)
1. Deep Learning is mainly based on:
a) Decision Trees
b) Artificial Neural Networks
c) Linear Regression
d) Clustering
Answer: b) Artificial Neural Networks
2. Which of the following is a type of Deep Learning model?
a) Convolutional Neural Network (CNN)
b) Support Vector Machine (SVM)
c) Decision Tree
d) K-Means
Answer: a) Convolutional Neural Network (CNN)
3. CNNs are mostly used for:
a) Text processing
b) Image recognition
c) Audio amplification
d) Sorting numbers
Answer: b) Image recognition
4. RNN stands for:
a) Random Neural Network
b) Recursive Neural Network
c) Recurrent Neural Network
d) Rapid Neural Network
Answer: c) Recurrent Neural Network
5. RNNs are best suited for:
a) Predicting house prices
b) Sequence and time-series data
c) Classifying static images
d) Sorting documents
Answer: b) Sequence and time-series data
6. LSTM is a special type of:
a) Convolutional Network
b) Recurrent Network
c) Generative Network
d) Feedforward Network
Answer: b) Recurrent Network
7. Which type of Deep Learning is best for Natural Language Processing (NLP)?
a) CNN
b) RNN / LSTM
c) Decision Trees
d) Naïve Bayes
Answer: b) RNN / LSTM
8. GANs are mainly used for:
a) Generating new data
b) Classifying documents
c) Sorting datasets
d) Compressing files
Answer: a) Generating new data
9. The two main parts of a GAN are:
a) Encoder and Decoder
b) Generator and Discriminator
c) Input and Output
d) Classifier and Regressor
Answer: b) Generator and Discriminator
10. Autoencoders are mainly used for:
a) Predicting time-series data
b) Data compression and feature extraction
c) Generating new images
d) Sorting datasets
Answer: b) Data compression and feature extraction
11. Which Deep Learning model is most commonly used in speech recognition?
a) CNN
b) RNN
c) GAN
d) Autoencoder
Answer: b) RNN
12. Which Deep Learning model is most effective in detecting faces in photos?
a) CNN
b) GAN
c) RNN
d) LSTM
Answer: a) CNN
13. Which Deep Learning model is best for detecting fraudulent financial transactions (time-sequence based)?
a) CNN
b) RNN
c) GAN
d) Autoencoder
Answer: b) RNN
14. GANs are an example of:
a) Supervised Learning
b) Unsupervised Learning
c) Reinforcement Learning
d) Hybrid Learning
Answer: b) Unsupervised Learning
15. Which Deep Learning model helps in dimensionality reduction?
a) CNN
b) RNN
c) Autoencoder
d) GAN
Answer: c) Autoencoder
16. Which of the following is an application of GANs?
a) Creating realistic fake images
b) Predicting stock prices
c) Classifying emails
d) Identifying spam
Answer: a) Creating realistic fake images
17. CNN layers mainly consist of:
a) Convolution, Pooling, Fully connected layers
b) Input, Output only
c) Only hidden layers
d) Random clustering
Answer: a) Convolution, Pooling, Fully connected layers
18. RNN suffers from which major problem?
a) Overfitting
b) Vanishing Gradient
c) Lack of features
d) Small dataset size
Answer: b) Vanishing Gradient
19. Which Deep Learning type is used for anomaly detection?
a) Autoencoder
b) GAN
c) CNN
d) RNN
Answer: a) Autoencoder
20. LSTM overcomes which RNN limitation?
a) Lack of labelled data
b) Vanishing gradient problem
c) Too many convolution layers
d) Limited dataset size
Answer: b) Vanishing gradient problem
Artificial Neural Networks
Artificial Neural Networks (20 with Answers)
1. What does ANN stand for in Artificial Intelligence?
a) Advanced Neural Node
b) Artificial Neural Network
c) Automated Numeric Network
d) Applied Neural Notation
Answer: b) Artificial Neural Network
2. ANN is inspired by which part of the human body?
a) Heart
b) Brain
c) Eye
d) Lungs
Answer: b) Brain
3. The basic unit of an ANN is called:
a) Neuron
b) Layer
c) Weight
d) Bias
Answer: a) Neuron
4. In ANN, the input signals are multiplied with:
a) Layers
b) Weights
c) Neurons
d) Biases
Answer: b) Weights
5. Which function decides whether a neuron should activate or not?
a) Loss function
b) Activation function
c) Cost function
d) Regression function
Answer: b) Activation function
6. Which of the following is NOT an activation function?
a) ReLU
b) Sigmoid
c) Softmax
d) Sorting
Answer: d) Sorting
7. ANN is made up of which three main layers?
a) Input, Hidden, Output
b) Input, Weight, Bias
c) Data, Model, Prediction
d) Training, Testing, Validation
Answer: a) Input, Hidden, Output
8. What is the main role of hidden layers in ANN?
a) Accepting raw input
b) Performing computations and learning patterns
c) Giving final output
d) Sorting input
Answer: b) Performing computations and learning patterns
9. Which of the following is true about ANN?
a) It always requires labelled data
b) It mimics the structure of the human brain
c) It does not use mathematics
d) It works without training
Answer: b) It mimics the structure of the human brain
10. What does the term “training ANN” mean?
a) Writing new code for neurons
b) Adjusting weights and biases to reduce error
c) Deleting hidden layers
d) Sorting input data
Answer: b) Adjusting weights and biases to reduce error
11. Which algorithm is commonly used to train ANNs?
a) Backpropagation
b) Sorting algorithm
c) Searching algorithm
d) Linear regression
Answer: a) Backpropagation
12. The process of comparing predicted output with actual output is called:
a) Forward propagation
b) Error calculation (Loss function)
c) Training
d) Feature scaling
Answer: b) Error calculation (Loss function)
13. ANN can be used in:
a) Image recognition
b) Speech recognition
c) Natural Language Processing
d) All of the above
Answer: d) All of the above
14. The mathematical model of a neuron is often called:
a) Perceptron
b) Transformer
c) Decoder
d) Generator
Answer: a) Perceptron
15. Which activation function outputs values between 0 and 1?
a) ReLU
b) Sigmoid
c) Tanh
d) Linear
Answer: b) Sigmoid
16. Overfitting in ANN happens when:
a) Model performs well on training but poorly on testing data
b) Model performs poorly on both training and testing data
c) Model has too few parameters
d) Training data is too small
Answer: a) Model performs well on training but poorly on testing data
17. Dropout is a technique used in ANN to:
a) Increase accuracy always
b) Reduce overfitting
c) Remove training data
d) Delete neurons permanently
Answer: b) Reduce overfitting
18. ANN learns patterns mainly through:
a) Rules and formulas
b) Experience and data
c) Manual programming
d) Fixed algorithms
Answer: b) Experience and data
19. Which of the following is NOT an application of ANN?
a) Medical diagnosis
b) Weather prediction
c) Image classification
d) Manual file storage
Answer: d) Manual file storage
20. What is the main limitation of ANN?
a) Can only work on images
b) Requires large amounts of data and computing power
c) Cannot process numeric data
d) Works without training
Answer: b) Requires large amounts of data and computing power