What is one purpose of Principal Component Analysis (PCA)?

Reduce features while preserving most of the data's variance.

In reinforcement learning, what does the agent learn to maximize?

A reward signal based on actions taken in an environment

What is the primary goal of feature selection?

To retain only the most relevant features for the target variable

What does data normalization do?

Scales numerical variables to have values between 0 and 1.

Why is data quality critical in machine learning projects?

High-quality data ensures better model interpretability and performance.

Which of the following is NOT an example of machine learning?

A manually programmed calculator for financial forecasting.

Which of the following best describes the goal of the “Evaluation” step in CRISP-DM?

Assess whether the model meets the business objectives

In CRISP-DM, during the deployment phase, which of the following is a typical activity?

Creating a system to integrate the model into the production environment

What is the main activity in the "Data Understanding" phase?

Summarize the data to detect patterns, outliers, and inconsistencies.

What is the main drawback of using one-hot encoding for categorical variables?

It increases the dimensionality of the dataset significantly.

What is the main difference between artificial intelligence (AI) and machine learning (ML)?

ML is a subfield of AI where machines learn from data

What is the primary objective of the “Business Understanding” step in CRISP-DM?

Translate the business problem into a data science problem.

Why is feature scaling necessary in some machine learning algorithms?

To ensure features contribute equally to the model.

studbud

Machine Learning Workshop ACM

15 termsStudBud Administration3 plays

Imported from external platform

Flashcards15 cardsFPractice quizPHost liveL

Questions 15

Answers

1
What is one purpose of Principal Component Analysis (PCA)?
Convert non-linear data into a linear form
Improve data interpretability by encoding categories as numbers.
Reduce features while preserving most of the data's variance.
Remove missing data from the dataset.
2
In reinforcement learning, what does the agent learn to maximize?
A reward signal based on actions taken in an environment
The number of actions taken
The loss function of the model
A predefined output for each input
3
Which of the following is a valid approach to handle missing data?
All of the above
Use algorithms that can handle missing data natively.
Drop rows or columns with missing values.
Replace missing values with statistical metrics like mean or median.
4
What is the primary goal of feature selection?
To select features that are highly correlated with each other.
To increase the size of the training dataset
To reduce the computational complexity of hyperparameter tuning
To retain only the most relevant features for the target variable
5
What does data normalization do?
Converts categorical variables into numerical values.
Scales numerical variables to have values between 0 and 1.
Removes redundant rows from the dataset.
Increases the overall dataset size
6
Why is data quality critical in machine learning projects?
High-quality data ensures better model interpretability and performance.
Poor data quality always leads to model underfitting
Data quality has no impact if the model is complex enough
Models can automatically correct data quality issues
7
Which of the following is NOT an example of machine learning?
A recommendation system that suggests products based on user behavior.
A manually programmed calculator for financial forecasting.
A program that detects spam emails using labeled examples.
A self-driving car that adapts its behavior over time.
8
How can you handle outliers in a dataset?
Transform them using log scaling
Replace them with the mean or median
All of the above
Remove them entirely from the dataset
9
Which of the following best describes the goal of the “Evaluation” step in CRISP-DM?
Generate predictions for unseen data.
Compare the performance of various machine learning algorithms
Optimize hyperparameters of the chosen model
Assess whether the model meets the business objectives
10
In CRISP-DM, during the deployment phase, which of the following is a typical activity?
Creating a system to integrate the model into the production environment
Finalizing the data cleaning pipeline.
Visualizing outliers and missing data.
Hyperparameter tuning for model optimization.
11
What is the main activity in the "Data Understanding" phase?
Develop visualizations for deployment dashboards.
Identify the data storage format for deployment.
Summarize the data to detect patterns, outliers, and inconsistencies.
Build a machine learning model to evaluate data quality.
12
What is the main drawback of using one-hot encoding for categorical variables?
It cannot handle variables with more than 10 categories.
It reduces the dataset size too much.
It introduces bias into the dataset.
It increases the dimensionality of the dataset significantly.
13
What is the main difference between artificial intelligence (AI) and machine learning (ML)?
A. ML is a separate field unrelated to AI
AI and ML are identical in practice
ML is a subfield of AI where machines learn from data
AI focuses on creating intelligent systems, while ML relies on programming
14
What is the primary objective of the “Business Understanding” step in CRISP-DM?
Collect and clean the dataset.
Translate the business problem into a data science problem.
Deploy the final model into production.
Build a baseline machine learning model.
15
Why is feature scaling necessary in some machine learning algorithms?
To reduce data redundancy.
To encode categorical variables
To ensure features contribute equally to the model.
To improve model interpretability

Machine Learning Workshop ACM Flashcards | StudBud