BADM 211 Test 3 Quizzes – Questions With
Proven Answers
Save
Terms in this set (46)
Suppose you are working binary classification problem
in the business loans
department for Bank X
and deciding on which
businesses to extend a
loan. To do this you want
to predict whether a
business applicant will go
bankrupt or remain
solvent in the next three
years. Therefore, this is a
regression problem
binary classification
problem
multi-class classification
problem
Most classifiers do not False
estimate propensities
representing the
probability that a record
belongs to each of the
classes.
True
False
,Suppose you are working T
in the business loans
department for Bank X
and deciding on which
businesses to extend a
loan. Business Y has
applied for a loan and
your model estimated the
probability of it being
bankrupt as 0.55 . If your
optimal decision
threshold is 0.6 then you
will approve the
application of Business Y.
True
False
Suppose for a record k, 0.2
the probability that it
belongs to the positive
class is 0.8 . Then the
probability it belongs to
the negative class is
0.4
0.2
0.6
1.6
, One way of determining maximizing overall accuracy
the optimal decision
threshold when both
classes are of equal
importance is by
minimizing overall
accuracy
maximizing overall error
maximizing overall
accuracy
minimizing sensitivity
Suppose you are False
predicting whether an
email is spam or not. In
your data set 98% of the
samples are not-spam.
Therefore, not-spam is
the minority class.
True
False
Suppose you are False
predicting you are
predicting which emails
are spam and which are
not-spam. In your data
set 98% of the samples
are not-spam. Therefore,
the naive rule is to assign
the class spam to each
email in the validation set.
True
False
Proven Answers
Save
Terms in this set (46)
Suppose you are working binary classification problem
in the business loans
department for Bank X
and deciding on which
businesses to extend a
loan. To do this you want
to predict whether a
business applicant will go
bankrupt or remain
solvent in the next three
years. Therefore, this is a
regression problem
binary classification
problem
multi-class classification
problem
Most classifiers do not False
estimate propensities
representing the
probability that a record
belongs to each of the
classes.
True
False
,Suppose you are working T
in the business loans
department for Bank X
and deciding on which
businesses to extend a
loan. Business Y has
applied for a loan and
your model estimated the
probability of it being
bankrupt as 0.55 . If your
optimal decision
threshold is 0.6 then you
will approve the
application of Business Y.
True
False
Suppose for a record k, 0.2
the probability that it
belongs to the positive
class is 0.8 . Then the
probability it belongs to
the negative class is
0.4
0.2
0.6
1.6
, One way of determining maximizing overall accuracy
the optimal decision
threshold when both
classes are of equal
importance is by
minimizing overall
accuracy
maximizing overall error
maximizing overall
accuracy
minimizing sensitivity
Suppose you are False
predicting whether an
email is spam or not. In
your data set 98% of the
samples are not-spam.
Therefore, not-spam is
the minority class.
True
False
Suppose you are False
predicting you are
predicting which emails
are spam and which are
not-spam. In your data
set 98% of the samples
are not-spam. Therefore,
the naive rule is to assign
the class spam to each
email in the validation set.
True
False
