Consider the following statements and choose the correct option: Assertion (S): The value of adjusted R2 is always less than the value of R2 Reason (R): Adjusted R2 accounts for the number of predictors in multiple linear regression model
Both S and R are true and R is the correct explanation of S
Both S and R are true but R is not the correct explanation of S
S is true but R is false
S is false but R is true
Both S and R are true and R is the correct explanation of S
If the structure of the data is complex and irregular, which of the following will be optimum value of ‘k’?
On the higher side
On the lower side
Equal to the total number of observations in the dataset
The value of ‘k’ has no impact
On the lower side
Which of the following statements is incorrect with respect to adjusted R-squared value?
Higher the number of predictors, higher the adjusted R-squared value
In a single predictor case, value of adjusted R-squared is equal to squared correlation
Adjusted R-squared uses a penalty on the number of predictors
Higher values of adjusted R-square indicate better fit
Higher the number of predictors, higher the adjusted R-squared value
Which among the following techniques can be used for variable selection and dimension reduction?
Exhaustive search
Partial iterative search
Both A and B
None of the above
Both A and B
In partial iterative search, which of the following algorithms are begun with the full model?
Forward selection
Backward selection
Exhaustive search
Stepwise regression
Backward selection
Which of the following algorithms overlooks the pairs or groups of predictors that perform well together but perform poorly as single predictors?
Forward selection
Backward selection
Exhaustive search
Both A and B
Forward selection
Compute the Euclidean distance for the following data points with 4 predictors: S(3,5,2,8) and T(1,4,6,2)
16.15
7.54
5
13
7.54
In k-NN technique, which of the following is a limitation of using high value of k? 
Fitting to local patterns
Dominance of global effect
Fitting to noise
None of the above
Dominance of global effect
Which of the following situations is regarded as a naïve rule in k-NN?
When k =1
When k >1
When k =n (where ‘n’ is the number of total observations)
When 1<k<n (where ‘n’ is the number of total observations)
When k =n (where ‘n’ is the number of total observations)
What changes are observed when k-NN is used for prediction task rather than classification task?
Computation of distance between the new observation and training partition records is different
Value of new record is determined using weighted average of all the k-nearest records
Value of new record is determined using weighted average of the records belonging to the dominant class
Overall misclassification error is used as performance metric
Value of new record is determined using weighted average of all the k-nearest records
