LEE-FELIR (F)

1. What is the defining characteristic of the training data used in supervised learning?

A) The data is generated randomly by the algorithm.
B) The data is generated randomly by the algorithm.
C) The data is unlabeled, and the model must find patterns on its own.
D) The @data is labeled, meaning each example is paired with a target output.

2. The primary goal of a supervised learning model is to:

A) Memorize the entire training dataset perfectly.
B) Discover hidden patterns without any guidance.
C) Reduce the dimensionality of the input data for visualization.
D) Gen@eralize from the training data to make accurate predictions on new, unseen data.

3. In the analogy of a child learning from flashcards, the animal's name on the card represents what component of supervised learning?

A) The@ label or target output.
B) The loss function.
C) The input features.
D) The model's parameters.

4. Which of the following tasks is a classic example of a classification problem?

A) Forecasting the temperature for tomorrow.
B) Predicting the selling price of a house based on its features.
C) Diagnosing @a tumor as malignant or benign based on medical images.
D) Estimating the annual revenue of a company.

5. A model that predicts the continuous value of a stock price for the next day is solving a:

A) Clustering problem.
B) Dimensionality reduction problem.
C) Regressio@n problem.
D) Classification problem.

6. What is the core objective of unsupervised learning?

A) To achieve perfect accuracy on a held-out test set.
B) To predict a target variable based on labeled examples.
C) To @discover the inherent structure, patterns, or relationships within unlabeled data.
D) To classify emails into spam and non-spam folders.

7. In the analogy of a child grouping toys without instructions, the act of putting all the cars together is most similar to which unsupervised learning technique?

A) Classification.
B) C@lustering.
C) Reinforcement Learning.
D) Regression.

8. Grouping customers based solely on their purchasing behavior, without pre-defined categories, is an application of:

A) A support vector machine for classification.
B) Linear Regression, a type of supervised learning.
C) Logistic Regression, a type of supervised learning.
D) Clu@stering, a type of unsupervised learning.

9. The main goal of dimensionality reduction techniques like PCA is to:

A) Increase the number of features to improve model accuracy.
B) Re@duce the number of features while preserving the most important information in the data.
C) Predict a continuous output variable.
D) Assign categorical labels to each data point.

10. Market basket analysis, which finds rules like "if chips then soda," is a classic example of:

A) Deep learning with neural networks.
B) Regression in supervised learning.
C) Ass@ociation rule learning in unsupervised learning.
D) Classification in supervised learning.

11. Semi-supervised learning is particularly useful in real-world scenarios because:

A) It requires no labeled data at all.
B) It is simpler to implement than unsupervised learning.
C) It is always more accurate than fully supervised learning.
D) La@beling data is often expensive and time-consuming, so it leverages a small labeled set with a large unlabeled set.

12. The fundamental question that a regression model aims to answer is:

A) "Which category?"
B) "Is this pattern anomalous?"
C) "What is the underlying group?"
D) "H@ow much?" or "How many?"

13. The fundamental question that a classification model aims to answer is:

A) "How can I reduce the number of features?"
B) "What is the correlation between these variables?"
C) "How much?" or "How many?"
D) "Whic@h category?" or "What class?"

14. Which algorithm is most directly designed for predicting a continuous target variable?

A) Lin@ear Regression
B) k-Nearest Neighbors for classification
C) Decision Tree for classification
D) Logistic Regression

15. A model that uses patient data to assign a label of "High," "Medium," or "Low" risk for a disease is performing:

A) Mult@i-class classification
B) Clustering
C) Regression
D) Dimensionality reduction

16. In a Decision Tree used for classification, what do the leaf nodes represent?

A) The input features for a new data point
B) The average value of a continuous target
C) The probability of moving to the next node
D) Th@e final class labels or decisions

17. In a Regression Tree, what is typically represented at the leaf nodes?

A) A c@ontinuous value, often the mean of the target values of the training instances that reach the leaf
B) A random number
C) The name of the feature used for splitting
D) A categorical class label

18. A key strength of Decision Trees is their:

A) Guarantee to find the global optimum for any dataset
B) Inter@pretability; the model's decision-making process is easy to understand and visualize
C) Immunity to overfitting on noisy datasets
D) Superior performance on all types of data compared to other algorithms

19. The "kernel trick" used in Support Vector Machines (SVMs) allows them to:

A) Perform linear regression more efficiently
B) Initialize the weights of a neural network
C) Grow a tree structure by making sequential decisions
D) Fin@d a linear separating hyperplane in a high-dimensional feature space, even when the data is not linearly separable in the original space

20. The "support vectors" in an SVM are the:

A) . Da@ta points that are closest to the decision boundary and most critical for defining the optimal hyperplane
B) The axes of the original feature space
C) All data points in the training set
D) The weights of a neural network layer

21. When comparing Decision Trees and SVMs, a primary advantage of SVMs is:

A) Their lower computational cost for very large datasets
B) Their superior interpretability and simplicity
C) Their inherent resistance to any form of overfitting
D) The@ir effectiveness in high-dimensional spaces and their ability to model complex, non-linear decision boundaries

22. The process in supervised learning where a model's parameters are adjusted to minimize the difference between its predictions and the true labels is called:

A) Clustering
B) Dimensionality reduction
C) Tr@aining or model fitting
D) Data preprocessing

23. A key challenge in unsupervised learning is evaluating model performance because:

A) There@ are no ground truth labels to compare the results against
B) The algorithms are not well-defined
C) The data is always too small
D) The models are always less accurate than supervised models

24. The task of reducing a 50-dimensional dataset to a 2-dimensional plot for visualization is best accomplished by:

A) An Association rule learning algorithm
B) Dimen@sionality Reduction techniques like Principal Component Analysis (PCA)
C) A Classification algorithm like Logistic Regression
D) A Regression algorithm like Linear Regression

25. If an e-commerce company wants to automatically group its products into categories without any pre-existing labels, it should use:

A) Classification, a supervised learning method
B) Clus@tering, an unsupervised learning method
C) A neural network for image recognition
D) Regression, a supervised learning method

26. The core building block of a neural network is a(n):

A) Artifi@cial neuron or perceptron, which receives inputs, applies a transformation, and produces an output
B) Support vector
C) Decision node in a tree
D) Principal component

27. In a neural network, the function inside a neuron that determines its output based on the weighted sum of its inputs is called the:

A) Activ@ation function
B) Loss function
C) Kernel function
D) Optimization algorithm

28. Which of the following is a non-linear activation function crucial for allowing neural networks to learn complex patterns?

A) A constant function
B) The mean squared error function
C) The identity function (f(x) = x)
D) Rectifie@d Linear Unit (ReLU)

29. The process of "training" a neural network involves:

A) Manually setting the weights based on expert knowledge
B) Iterativ@ely adjusting the weights and biases to minimize a loss function
C) Randomly assigning weights and never changing them
D) Clustering the input data

30. Backpropagation is the algorithm used in neural networks to:

A) Perform clustering on the output layer
B) Visualize the network's architecture
C) Efficient@ly calculate the gradient of the loss function with respect to all the weights in the network, enabling the use of gradient descent
D) Initialize the weights before training

31. Deep Learning is a subfield of machine learning that primarily uses:

A) Neural n@etworks with many layers (hence "deep")
B) K-means clustering exclusively
C) Decision trees with a single split
D) Simple linear regression models

32. A key advantage of deep neural networks over shallower models is their ability to:

A) Auto@matically learn hi@erarchical feature representations from data
B) Always train faster and with less data
C) Be perfectly interpretable, like a decision tree
D) Operate without any need for data preprocessing

33. Convolutional Neural Networks (CNNs) are particularly well-suited for tasks involving:

A) Unsupervised clustering of audio signals
B) Text data and natural language processing
C) Tabular data with many categorical features
D) Image @data, due to their architecture which exploits spatial locality

34. The "convolution" operation in a CNN is designed to:

A) Perform the final classification
B) Initialize the weights of the network
C) Flatten the input into a single vector
D) Detect@ local features (like edges or textures) in the input by applying a set of learnable filters

35. Recurrent Neural Networks (RNNs) are designed to handle:

A) Independent and identically distributed (IID) data points
B) Static, non-temporal data
C) Only image data
D) Sequ@ential data, like time series or text, due to their internal "memory" of previous inputs

36. The "vanishing gradient" problem in deep networks refers to:

A) The@ gradients becoming exceedingly small as they are backpropagated through many layers, which can halt learning in early layers
B) The model overfitting to the training data
C) The gradients becoming too large and causing numerical instability
D) The loss function reaching a perfect value of zero

37. The "training set" is used to:

A) Tune the model's hyperparameters
B) Deploy the model in a production environment
C) Provide an unbiased evaluation of a final model's performance
D) Fit th@e model's parameters (e.g., the weights in a neural network)

38. The "validation set" is primarily used for:

A) Tun@ing hyperparameters and making decisions about the model architecture during development
B) The initial training of the model's weights
C) Data preprocessing and cleaning
D) The final, unbiased assessment of the model's generalization error

39. The "test set" should be:

A) Use@d only once, for a final evaluation of the model's performance on unseen data after model development is complete
B) Used repeatedly to tune the model's hyperparameters
C) Used repeatedly to tune the model's hyperparameters
D) Ignored in the machine learning pipeline

40. Overfitting occurs when a model:

A) Learns @the training data too well, including its noise and outliers, and performs poorly on new, unseen data
B) Is evaluated using the training set instead of a test set
C) Fails to learn the underlying pattern in the training data
D) Is too simple to capture the trends in the data

41. A common technique to reduce overfitting in neural networks is:

A) Training for more epochs without any checks
B) Increasing the model's capacity by adding more layers
C) Dropo@ut, which randomly ignores a subset of neurons during training
D) Using a smaller training dataset

42. The "bias" of a model refers to:

A) The activation function used in the output layer
B) The error from sensitivity to small fluctuations in the training set, leading to overfitting
C) The erro@r from erroneous assumptions in the learning algorithm, leading to underfitting
D) The weights connecting the input layer to the hidden layer

43. The "variance" of a model refers to:

A) The speed at which the model trains
B) The er@ror from sensitivity to small fluctuations in the training set, leading to overfitting
C) The error from erroneous assumptions in the learning algorithm, leading to underfitting
D) The intercept term in a linear regression model

44. The "bias-variance tradeoff" implies that:

A) Only variance is important for model performance
B) Only bias is important for model performance
C) Decrea@sing bias will typically increase variance, and vice versa. The goal is to find a balance
D) Bias and variance can be minimized to zero simultaneously

45. A learning curve that shows high training accuracy but low validation accuracy is a classic sign of:

A) Perfect model performance
B) Overf@itting
C) Underfitting
D) A well-generalized model

46. In a neural network, the "loss function" (or cost function) measures:

A) The speed of the backpropagation algorithm
B) The number of layers in the network
C) The accuracy on the test set
D) How well the model is performing on the training data; it's the quantity we want to minimize during training

47. Gradient Descent is an optimization algorithm that:

A) Iteratively adjusts parameters in the direction that reduces the loss function
B) Guarantees finding the global minimum for any loss function
C) Is only used for unsupervised learning
D) Randomly searches the parameter space for a good solution

48. The "learning rate" in gradient descent controls:

A) The amount of training data used in each epoch
B) The activation function for the output layer
C) The size of the step taken during each parameter update. A rate that is too high can cause divergence, while one that is too low can make training slow
D) The number of layers in a neural network

49. "Epoch" in neural network training refers to:

A) One complete pass of the entire training dataset through the learning algorithm
B) The processing of a single training example
C) A type of regularization technique
D) The final evaluation on the test set

50. "Batch Size" in neural network training refers to:

A) The number of validation examples
B) The number of layers in the network
C) The total number of examples in the training set
D) The number of training examples used in one forward/backward pass before the model's parameters are updated

51. "Epoch" in neural network training refers to:

A) One complete pass of the entire training dataset through the learning algorithm
B) The processing of a single training example

52. "Epoch" in neural network training refers to: