Data Preprocessing and ML basics Quiz

Which metric is more important when the cost of a false positive is very high?

• Precision

• Recall

• F1 score

• Accuracy

Which type of machine learning is used for classifying emails as spam or not spam?

• Semi-supervised learning

• Supervised learning

• Unsupervised learning

• Reinforcement learning

Which of the following is not a technique used in unsupervised learning?

• K-means clustering

• Principal Component Analysis (PCA)

• Decision Trees

• Hierarchical clustering

Which type of machine learning algorithm is commonly used for anomaly detection?

• Supervised learning

• Unsupervised learning

• Reinforcement learning

• Semi-supervised learning

What is the range of values that the sigmoid function outputs?

• -1 to 1

• 0 to 1

• 0 to infinity

• Negative infinity to positive infinity

What is the objective of the cost function in linear regression?

• To find the most complex model that fits the data

• To reduce the dimensionality of the data

• To maximize the number of correct predictions

• To minimize the sum of squared differences between actual and predicted values

What does a boxplot display that helps in identifying outliers?

• Mean and standard deviation

• Median and interquartile range

• Histogram of the data

• Frequency distribution

Why might one-hot encoding be problematic with high cardinality categorical features?

• It requires the data to be normalized first

• It can introduce multicollinearity

• It increases the dimensionality of the dataset significantly

• It is only suitable for ordinal data

Which encoding method involves replacing the categorical values with a number representing the mean of the target variable for that category?

• Dummy encoding

• Target encoding

• One-hot encoding

• Label encoding

When would you use standardization over normalization?

• When the data does not follow a Gaussian distribution

• When the data follows a Gaussian distribution and outliers are not a concern

• When outliers are present

• When the model requires the data to be in a specific range