This set of questions covers a range of foundational topics in AI and ML, including supervised and unsupervised learning, model evaluation, and applications in natural language processing and deep learning. These questions are designed to test knowledge across various core concepts in AI/ML, suitable for beginners and those preparing for certification exams in these fields.
To Check your Progress and get points for your performance
Login & Start the Test
OR
Start the Test Without Login
OR
Scroll Down to See Questions
Direction: Supervised learning involves training a model on labeled data.
Q.1.
What is the main objective of supervised learning in machine learning?
- To predict outcomes for new data
- To discover hidden patterns in data
- To reduce the dimensionality of data
- To generate new data from existing data
- None of the above
Answer:
(A)
Explanation: In supervised learning, the model is trained to predict outcomes based on labeled input data.
Explanation: In supervised learning, the model is trained to predict outcomes based on labeled input data.
Direction: Classification tasks involve categorizing data into predefined classes.
Q.2.
Which algorithm is commonly used for classification tasks?
- K-Means Clustering
- Linear Regression
- Decision Trees
- Apriori Algorithm
- DBSCAN
Answer:
(C)
Explanation: Decision Trees are widely used for classification tasks due to their interpretability and ability to handle both numerical and categorical data.
Explanation: Decision Trees are widely used for classification tasks due to their interpretability and ability to handle both numerical and categorical data.
Direction: Overfitting is a common problem when a model learns too much from the training data.
Q.3.
What does "overfitting" refer to in machine learning?
- A model performing well on both training and test data
- A model performing poorly on training data but well on test data
- A model performing well on training data but poorly on test data
- A model that generalizes well to unseen data
- A model that performs equally on all datasets
Answer:
(C)
Explanation: Overfitting occurs when a model is too complex and captures noise in the training data, leading to poor generalization on new data.
Explanation: Overfitting occurs when a model is too complex and captures noise in the training data, leading to poor generalization on new data.
Direction: Unsupervised learning deals with unlabeled data to find patterns.
Q.4.
Which of the following is an example of unsupervised learning?
- Support Vector Machines
- Principal Component Analysis (PCA)
- Random Forest
- Naive Bayes
- Linear Discriminant Analysis
Answer:
(B)
Explanation: PCA is an unsupervised learning method used for dimensionality reduction and to identify patterns in data without labels.
Explanation: PCA is an unsupervised learning method used for dimensionality reduction and to identify patterns in data without labels.
Direction: Learning rate is a key hyperparameter in training neural networks and other models.
Q.5.
What is the purpose of a "learning rate" in gradient descent optimization?
- To determine the size of steps taken towards the minimum of a function
- To increase the model's complexity
- To define the number of iterations
- To prevent the model from underfitting
- To ensure the model converges to a local maximum
Answer:
(A)
Explanation: The learning rate controls how much to change the model in response to the estimated error each time the model weights are updated.
Explanation: The learning rate controls how much to change the model in response to the estimated error each time the model weights are updated.
Direction: Tokenization is a fundamental step in text preprocessing.
Q.6.
In natural language processing, what does "tokenization" refer to?
- Breaking down text into paragraphs
- Splitting text into sentences
- Breaking text into smaller units like words or phrases
- Removing punctuation and stopwords
- Translating text into another language
Answer:
(C)
Explanation: Tokenization involves splitting text into smaller units, such as words or phrases, which can then be analyzed and processed.
Explanation: Tokenization involves splitting text into smaller units, such as words or phrases, which can then be analyzed and processed.
Direction: Evaluation metrics help measure the effectiveness of a model's predictions.
Q.7.
Which metric is commonly used to evaluate the performance of a classification model?
- Mean Squared Error (MSE)
- R-squared
- F1 Score
- Root Mean Squared Error (RMSE)
- Adjusted R-squared
Answer:
(C)
Explanation: The F1 Score is commonly used for evaluating classification models, especially when dealing with imbalanced datasets, as it considers both precision and recall.
Explanation: The F1 Score is commonly used for evaluating classification models, especially when dealing with imbalanced datasets, as it considers both precision and recall.
Direction: CNNs are a type of deep learning model particularly effective in certain applications.
Q.8.
What is a "convolutional neural network" (CNN) primarily used for?
- Time series forecasting
- Text generation
- Image recognition
- Speech synthesis
- Reinforcement learning
Answer:
(C)
Explanation: CNNs are specifically designed for processing structured grid data, like images, and are widely used in image and video recognition tasks.
Explanation: CNNs are specifically designed for processing structured grid data, like images, and are widely used in image and video recognition tasks.
Direction: An agent interacts with an environment to learn optimal actions.
Q.9.
In reinforcement learning, what is an "agent"?
- The environment where actions are taken
- A set of possible actions
- A reward function
- An entity that learns and makes decisions
- A policy for decision-making
Answer:
(D)
Explanation: In reinforcement learning, an agent is an entity that takes actions based on its observations of the environment to maximize cumulative reward.
Explanation: In reinforcement learning, an agent is an entity that takes actions based on its observations of the environment to maximize cumulative reward.
Direction: Regularization techniques help in preventing overfitting.
Q.10.
What is "regularization" in machine learning?
- Increasing model complexity
- Adding noise to the data
- Penalizing complex models
- Optimizing hyperparameters
- Reducing model accuracy
Answer:
(C)
Explanation: Regularization adds a penalty for more complex models to prevent overfitting, encouraging simpler models that generalize better.
Explanation: Regularization adds a penalty for more complex models to prevent overfitting, encouraging simpler models that generalize better.
SIMILAR FOR YOU
Videos from Youtube 〉
Posted in ai and ml machine learning artificial intelligence quiz on ai quiz on machine learning
103 views
Share