Practice multiple choice questions on Regression with answers. This is one of the fundamental techniques in Machine Learning which is widely used in basic problems.
If you want to revise the concept, read this article 👉:
Let us start with the questions. Click on the right option and the answer will be explained.
Question 1
What is Regression?
It is a technique to predict values
It is a technique to fix data
It is a Machine Learning algorithm
It is a technique to find outliers
Regression is used to create a relationship between a dependent variable to a one or more independent variables.
Question 2
What is a dependent variable
The value we want to predict
The features of our dataset
The parameters of the regression algorithm
The values that interfere in the value we want to predict
Dependent variable is the value we want to predict. Imagine that we want to know how much is the value of a house. It is the dependent variable and we have to consider it´s size, neighborhood, how many rooms, how many bathroom, does it have a garden, among other variables. It´s value depends of all these informations, and that is why it is called Dependent Variable.
Question 3
What are independent variables?
The values that interfere in the value we want to predict
The features of our dataset
The parameters of the regression algorithm
The value we want to predict
Following the house price example, all variables that can interfere in the house price can be called as independent variable. House size, neighborhood, how many rooms, when it was built, etc.
Question 4
What are outliers?
Extreme datapoints in our dataset
Values that are correlated to eachother
It is the main trend of our dataset
It is a regression technique
Outliers are extreme datapoints in our dataset that have too much more or less value than other datapoints. Most of the times outliers can be excluded from the dataset in order to preserve the regression quality. On the other hand, if we are working to prevent fraud, outliers is what we will be looking for, since their represent suspicious behavior.
Question 5
What is Multicollinearity?
High correlation between independent variables
Low correlation between independent variables
Correlation between outliers
Correlation between features
When we have 2 or more independent variables with high correlation, we call it of Multicollinearity. It can be harmful to our regression because make harder to ranking the variables in order to know which one interfere more in our dependent variable. In these cases, we usually keep only one of those variables and discard the others.
Question 6
What is overfitting?
Great result in training and poor result in test
Great result in training and great result in test
Poor result in training and poor result in test
Poor result in training and poor result in test
When we use unnecessary explanatory variables it might lead to overfitting.
Question 7
What are Linear and Logistic regression?
There are types of regression
It is how you can classify a regression
A regression must be Linear or Logistic
There are types of overfitting
Although they are the most known types of Regression, there are many others.
Q8. Linear Regression is an example of?
a) Supervised Learning
b) Unsupervised Learning
c) Semi-Supervised Learning
Answer: a) Supervised Learning
Linear Regression is an example of Supervised Learning. In fact, as a rule, all Regression techniques are an example of Supervised Learning.
Q9. Logistic Regression is an example of?
a) Supervised Learning: Regression
b) Classification
b) Unsupervised Learning
c) Semi-Supervised Learning
Answer: b) Classification
Logistic Regression is an example of Classification which is a type of Supervised Learning. As mentioned previously, all Regression techniques are an example of Supervised Learning. Exception is that Logistic Regression is not counted as a regression technique but as a Classification technique.
Question 10
Which answer explains better Linear Regression?
Dependent variable is continuous, independent variable(s) can be continuous or discrete, and nature of regression line is linear.
Dependent variable is discrete, independent variable(s) can be continuous or discrete, and nature of regression line is linear.
Dependent variable is continuous, independent variable(s) can be continuous or discrete, and nature of regression line is non-linear.
Dependent variable is discrete, independent variable(s) can be continuous or discrete, and nature of regression line is non-linear.
Linear Regression creates a relation between the dependent variables and all the independent variables using a best fit straight line, known as regression line.
Question 11
When is appropriate to use Logistic Regression?
When the dependent variable is binary
When the independent variables are binary
When the dependent variable is not binary
When the independent variables are not binary
Logistic Regression is used when we are looking for a binary value. This type of regression calculates the probability of a event has success or failure. It is widely used for classification.
Question 12
For what Polynomial Regression is used?
Handle with non-linear and separable data
Handle linear and separable data
Classify binary data
Find the best linear line
When we handle with non-linear and separable data, a straight line will not work. In this case, Polynomial is very usefull and their always have independents variables with power higher than 1.
Question 13
When we use Ridge Regression?
When our data have multicollinearity
When our data doens`t have multicollinearity
When we have a lot of outliers
When there is no outlier
Ridge Regression have a regularization parameter to fix the multicollinearity problem. It shrinks the value of coefficients but doesn’t reaches zero, which suggests no feature selection feature.
Q14. Which regression is used in the following image?
a) Linear Regression
b) Logistic Regression
c) Polynomial Regression
d) Ridge Regression
Answer: c) Polynomial Regression
Multiple curves in a line denote the graph is of a polynomial of multiple degree and hence, it is using Polynomial Regression.
Question 15
Which way Lasso Regression differs from Ridge Regression?
It uses absolute values in regularization parameter, instead of squares
It uses square values in regularization parameter
It works better in small datasets
It works better in big datasets
Lasso Regression uses absolute values in the penalty function, instead of squares. The result is penalize values which causes some of the parameter estimates to turn out excactly zero.
Question 16
What is ElasticNet Regression?
It is a mix of Lasso and Ridge Regression
It is the newest type of regression
It is the best way to use regression in Machine Learning
It is a type of regression focused in outliers
It is a combination of L1 and L2 regularization.
Q17. Consider the following image. Which one is used for Low Bias Low Variance?
a) Ridge Regression
b) Lasso Regression
c) Elastic Net Regression
d) Linear Regression
Answer: c) Elastic Net Regression
Ridge and Lasso Regression is used for high bias and high variance.
The scenario we are looking for is with Low Bias and Low Variance in order to have a better prediction from our model. Then we use regularization to reduce our variance and introducing some Bias.
We already did that using Ridge and Lasso, but both of them has faults. Elastic Net was created to combine the penalties of ridge regression and lasso to get the best of both worlds. Elastic Net aims at minimizing the loss information.
Q18. Which Regression technique uses F-test or T-test?
a) Ridge Regression
b) Stepwise Regression
c) Elastic Net Regression
d) Linear Regression
Answer: b) Stepwise Regression
Stepwise regression is a technique which adds or removes variables via series of F-tests or T-tests. The variables to be added or removed are chosen based on the test statistics of the estimated coefficients.
With these questions on Regression at OpenGenus, you must have a good idea of Regression. Enjoy.