BADM 211 BUNDLED EXAMS COMPLETE STUDY
GUIDE WITH ANSWERS
◉ Why would we partition our data into training and validation
partitions?
To test the model performance on data that has not been used to
develop the model
Because the data could be too large to work with
To fit different models on the two datasets and take average
performance of the two models
To fit separate models on each partition and compare models for
their performance Answer: To test the model performance on data
that has not been used to develop the model
◉ Outcome variables are relevant in the case of _________.
c) Classification
Both B and C
a) Clustering
b) Prediction Answer: Both B and C
,◉ Which of the following is an example of supervised learning?
A seasoned data scientist attempting to develop a new algorithm to
create multidimensional charts.
A store manager, under the supervision of her manager, attempting
to group customers into a few segments.
A newbie data scientist tasked with reducing the number of
predictors based on the correlation among them.
A store manager attempting to predict sales for tomorrow using a
model. Answer: A store manager attempting to predict sales for
tomorrow using a model.
◉ Which of the following could be a likely cause of overfitting?
Too many observations in the data
Many predictors are uncorrelated
Almost none of the predictors are correlated to each other
Too many predictors Answer: Too many predictors
◉ Which of the following is always true with respect to outliers?
a. Outliers are the values that are over three standard deviation
away from the mean
b. Outliers are invalid data points Answer: Only A
, ◉ Classification and prediction are two types of UNsupervised
learning.
False, in both the problems an outcome variable is specified
True, neither has an outcome variable specified
True, in both the problems an outcome variable is specified
False, neither has an outcome variable specified Answer: False, in
both the problems an outcome variable is specified
◉ Which statement is true regarding prediction and classification
problems?
Predicting whether it would rain tomorrow is an example of
classification problem.
The same problem may be both prediction and classification at the
same time.
Predicting whether it would rain tomorrow is an example of
prediction problem.
Prediction is concerned with predicting a categorical outcome while
classification is concerned with grouping similar observations
together. Answer: Predicting whether it would rain tomorrow is an
example of classification problem.
GUIDE WITH ANSWERS
◉ Why would we partition our data into training and validation
partitions?
To test the model performance on data that has not been used to
develop the model
Because the data could be too large to work with
To fit different models on the two datasets and take average
performance of the two models
To fit separate models on each partition and compare models for
their performance Answer: To test the model performance on data
that has not been used to develop the model
◉ Outcome variables are relevant in the case of _________.
c) Classification
Both B and C
a) Clustering
b) Prediction Answer: Both B and C
,◉ Which of the following is an example of supervised learning?
A seasoned data scientist attempting to develop a new algorithm to
create multidimensional charts.
A store manager, under the supervision of her manager, attempting
to group customers into a few segments.
A newbie data scientist tasked with reducing the number of
predictors based on the correlation among them.
A store manager attempting to predict sales for tomorrow using a
model. Answer: A store manager attempting to predict sales for
tomorrow using a model.
◉ Which of the following could be a likely cause of overfitting?
Too many observations in the data
Many predictors are uncorrelated
Almost none of the predictors are correlated to each other
Too many predictors Answer: Too many predictors
◉ Which of the following is always true with respect to outliers?
a. Outliers are the values that are over three standard deviation
away from the mean
b. Outliers are invalid data points Answer: Only A
, ◉ Classification and prediction are two types of UNsupervised
learning.
False, in both the problems an outcome variable is specified
True, neither has an outcome variable specified
True, in both the problems an outcome variable is specified
False, neither has an outcome variable specified Answer: False, in
both the problems an outcome variable is specified
◉ Which statement is true regarding prediction and classification
problems?
Predicting whether it would rain tomorrow is an example of
classification problem.
The same problem may be both prediction and classification at the
same time.
Predicting whether it would rain tomorrow is an example of
prediction problem.
Prediction is concerned with predicting a categorical outcome while
classification is concerned with grouping similar observations
together. Answer: Predicting whether it would rain tomorrow is an
example of classification problem.
