1. What is/are the outcome of optimizing the likelihood function ?
View Answer
Parameters
2. ________________ Regression can be used to model the data when the outcome is dichotomous .
View Answer
Logistic
3. When we normally multiply the coefficients with covariates we get a numeric value ? But when we need to get an output in the range 0 to 1 what do we need to perform ?
View Answer
Non Linear Transformation
4. A linear Model best predicts when the dependent variable is Binary.
View Answer
False
5. The systematic component in a GLM consists of Linear Predictor . What are they ?
View Answer
Covariate and Coefficient
6. Linear Models are best used when there are constraints that the outcome has to be strictly positive.
View Answer
False
7. (True Negative) / (True Negative + False Positive ) is what ?
View Answer
Specificity
8. What is the main objective of Bagging ?
View Answer
Variance Reduction by Aggregation
9. Bagging and Boosting have a very low memory footprint.
View Answer
False
10. What is the link function in a Poisson Regression model ?
View Answer
Log
11. Classification and Regression Trees are a set of ______________ Algorithms
View Answer
Non Linear Learning
12. Bagging is and abbreviation for _________________ Bootstrap
View Answer
Aggregation
13. __________ and ________ are techniques that are used for combining multiple models to improve overall accuracy.
View Answer
Bagging and Boosting
14. What is the main objective of Boosting ?
View Answer
Reduce Prediction Bias
15. What is the output of the Bayesian Regression Model ?
View Answer
Output along with the Probability Distribution
16. Bayesian Regression is Robust to Gaussian Noise
View Answer
True
17. What is the principal concept used in fitting a Tree Algorithm ?
View Answer
Variance Reduction
18. What class of algorithms do Trees belong ?
View Answer
Greedy Algorithms
19. Bagging and Boosting support Stochastic Learning
View Answer
True
20. In poisson regression the dependent variable represents _____________ or ______________
View Answer
Count or Rate
21. In mathematical terms Poisson Regression models _____________ of the count data .
View Answer
Logarithm
22. ROC curve is used to gauge the cost of raising a false alarm
View Answer
True
23. In _____ , modeling is done on the scale in which the data is captured.
View Answer
Generalized Linear Models
24. (True Positive) / (True Positive + False Negative ) is what ?
View Answer
Sensitivity
25. ROC stands for
View Answer
Receiver Operator Characteristics
26. Mean Standard Error is a good evaluation measure for Ranking Problem .
View Answer
False
27. In a confusion matrix what do the values in the Row signify ?
View Answer
Actual Outcome
28. What are the set of Go to algorithms to model non-linear behavior ?
View Answer
Stochastic Regression
No comments:
Post a Comment