CDSP 4 Quiz

1. What is the primary goal of machine learning in data science?

A: To replace human decision-making
B: To continually improve decisions based on data
C: To simulate natural intelligence in robots
D: To process data without the need for human supervision

2. What is the key difference between machine learning and deep learning?

A: Machine learning requires no data for training
B: Deep learning works with unstructured data, while machine learning generally uses structured data
C: Machine learning models are always more accurate than deep learning models
D: Deep learning requires less computational power than machine learning

3. Which machine learning algorithm is best suited for classification problems?

A: Linear regression
B: k-means clustering
C: Decision trees
D: ARIMA

4. What does the term “training data” refer to in machine learning?

A: Data used to test the performance of a model
B: Data that the algorithm uses to learn and create the model
C: Data collected from the final predictions of a model
D: Data that has been cleaned and processed

5. What is a key benefit of cross-validation in machine learning?

A: It improves the complexity of the model
B: It ensures the model is not overfitting to the training data
C: It speeds up the model’s training process
D: It reduces the need for labeled data

6. What is the primary risk of data leakage in machine learning?

A: The model underperforms on the training data
B: The test data influences the training process
C: The model becomes too complex to interpret
D: The model is unable to detect outliers in the data

7. What is a hyperparameter in machine learning?

A: A parameter learned from the training data
B: A value set by the machine learning practitioner before training
C: A variable that changes automatically during model training
D: A setting that improves data preprocessing

8. Which technique is typically used to handle imbalanced datasets?

A: SMOTE
B: Random undersampling
C: Feature scaling
D: Principal Component Analysis (PCA)

9. What is the purpose of a confusion matrix in machine learning?

A: To calculate the accuracy of a model
B: To compare the true labels with predicted labels
C: To optimize hyperparameters
D: To balance the dataset

10. What does the bias-variance tradeoff in machine learning refer to?

A: Balancing the complexity of the model
B: The tradeoff between data size and accuracy
C: The choice between underfitting and overfitting
D: The method of selecting the training algorithm

11. What is underfitting in machine learning?

A: A model that is too complex for the data
B: A model that fails to capture the underlying pattern in the data
C: A model with high variance
D: A model with high accuracy but low precision

12. What is an example of an activation function used in neural networks?

A: k-means
B: ReLU
C: Decision trees
D: Linear regression

13. What is one key advantage of using deep learning for image recognition tasks?

A: It requires minimal computational power
B: It performs well with structured data
C: It can automatically learn and extract features from raw data
D: It has limited scalability

14. What does the term “learning rate” refer to in machine learning?

A: The speed at which data is processed
B: The rate at which a model’s performance improves
C: The size of the steps taken when updating model weights
D: The amount of data fed into the model during each iteration

15. What is the purpose of backpropagation in neural networks?

A: To initialize network weights
B: To optimize the network structure
C: To update the weights based on the error in predictions
D: To normalize the input data

16. Which machine learning technique is commonly used for natural language processing (NLP) tasks?

A: Logistic regression
B: Convolutional neural networks (CNNs)
C: Recurrent neural networks (RNNs)
D: Random forest

17. What is an artificial neural network (ANN) designed to approximate?

A: A decision tree
B: A biological neural network
C: A regression model
D: A support vector machine

18. What is one primary function of the sigmoid activation function in neural networks?

A: It outputs values between -1 and 1
B: It is used to classify between more than two classes
C: It outputs values between 0 and 1 for binary classification
D: It normalizes the input data

19. What is one key characteristic of supervised learning?

A: It works without labeled data
B: It requires labeled data to train the model
C: It generates new labels for the data
D: It clusters data into different groups

20. What is one advantage of using decision trees over other classification algorithms?

A: They require large amounts of data to perform well
B: They are easy to interpret and visualize
C: They always result in the most accurate model
D: They automatically prevent overfitting