Quiz Topic - Regression

1. Suppose you have to predict the salary of an employee from their years of experience where the dataset has a salary range from 10000 to 50000. In which of the intervals your regressive model should predict?

view answer: D. 10000 to 50000
2. In simple linear regression, if you change the input value by 1 then output value will be changed by:

view answer: B. The slope parameter
3. You can compute the residual by-

view answer: A. actual y‐coordinate value - predicted y‐coordinate value
4. How to see the value of residuals geometrically

view answer: A. The perpendicular distance between a data point and the regression line
5. The equation of the regression line is y = 5x + 3. Predict y when x = 8.

view answer: A. 43
6. The equation of the regression line is y = 8x - 2. Compute the residual for the point (4, 28)

view answer: C. -2
7. What would be the best regression model for more than one independent variable?

view answer: B. Multiple Linear Regression
8. Suppose you have observed that you data has an exponential growth tendency. Then what regression model you should use-

view answer: C. Polynomial regression
9. Can we perform linear regression with a neural network?

view answer: A. Yes, we can
10. If you get a poor accuracy using a simple linear regression model. What will be the cause behind it-

view answer: C. Both A or B depending on the context
11. If your data grows in a non-linear fashion. Which model won’t perform well?

view answer: C. Simple linear regression
12. Suppose you got a training accuracy of 90% and a test accuracy of 50%. What happened with your model-

view answer: A. The model was over fitted with the training data
13. What is a support vector?

view answer: C. The distance between two boundary data points
14. What is a kernel?

view answer: B. A function that maps the value from one dimension to the other
15. Which of the following is not a kernel?

view answer: D. None
16. What does epsilon represent in Support Vector Regression?

view answer: B. Error threshold
17. In Regression, a decision tree splits the dataset based on-

view answer: A. Information entropy
18. Which one is a different algorithm?

view answer: A. Logistic Regression
19. Which one is not a better algorithm in the sense of overfitting?

view answer: A. Simple linear regression
20. If the actual value of a data point is 50 and the predicted value is 55, what will be the Mean Absolute Error(MAE)

view answer: B. 5
21. Which of the following is a regression algorithm?

view answer: A. Linear Regression
22. Suppose you have to predict the salary of employees from their experience. This is a-

view answer: B. Regression task
23. Regression is a-

view answer: A. Supervised Learning Algorithm
24. Which of the following is/are true about Normal Equation?

view answer: D. All of them
25. Which methods are used to find the best fit line in linear regression?

view answer: D. Least Square Error
26. What will happen when you increase the size of training data?

view answer: C. Bias increases and Variance decreases
27. If you fit 2 degree polynomial in linear regression-

view answer: B. The model will underfit the data
Explanation: Higher degree polynomials have chances to underfit at a lower degree.
28. Which of the following evaluation metrics can be used for Regression?

view answer: B. Mean-Squared-Error
Explanation: Regression gives continuous output. So, we use Mean-Squared-Error or MSE as evaluation metric. Rest are used in classification.
29. Linear regression is-

view answer: A. sensitive to outliers
Explanation: The regression line changes due to outliers. So, it is sensitive to outliers.
30. What is true about Residuals?

view answer: B. Lower is better

© aionlinecourse.com All rights reserved.