BSNS 112 Qualitative Data & Non-linear
and Computational Modelling Actual
Exam| 74 Real Questions With Detailed
Answers| Guaranteed Pass
1. What is the 10-way Cross Validation method used to do? - ANSWER✔️-
Measure models predictive accuracy and recognize how well a model will
predict data it has not seen.
2. How does the 10-way Cross Validation method work? - ANSWER✔️-
Divide data into 10 parts, train on 9, test on 1
3. What are decision trees split on? - ANSWER✔️-The best attribute to split
4. Does the accuracy of a decision tree increase as depth does? - ANSWER✔️-
Yes if same data is used for training and testing, no if cross validation test
used.
5. Can a decision tree handle both qualitative and quantitative data? -
ANSWER✔️-Yes
6. What are the three layers in an Artificial Neural Network? - ANSWER✔️-
Input Layer, Output Layer, Hidden Layer
7. What does Backpropagation in an ANN do? - ANSWER✔️-Begins with an
error output, propagates the error backward to calculate yield output error
values for neurons.
Then adjusts weights to better predict output from the same inputs.
Only applies to feed forward ANNs.
8. Values of x and y for a model can be referred to as? - ANSWER✔️-Input,
independent, dependent, output, explanatory, response
, 9. Can you run categorical variables through a linear model? - ANSWER✔️-
Yes, but can only give a binary outcome.
10. MLR with categorical variables yes (1) and no (0), what outcome gives yes?
- ANSWER✔️-Anything above 0
11.What transformations can be used on a linear regression model? -
ANSWER✔️-Log normal, cubic, square root, inverse squared
12.When fitting a model, R^2 should be... - ANSWER✔️-High
13.What does Backpropagation in an ANN NOT do? - ANSWER✔️-Makes the
hidden Layer visible.
14.What is the optimal number for hidden nodes for ANN? - ANSWER✔️-2-4
15.Why is scaling important? - ANSWER✔️-Pick clusters based on scale, so x
and y axis scales need to be constant in order to pick optimal cluster.
16.What is not true about scaling? - ANSWER✔️-Because scaling is in terms
of dimensions, it doesn't matter
17.Why is scaling done? - ANSWER✔️-To avoid bias
18.How are values normally scaled? - ANSWER✔️-To give a mean of 0 and
standard deviation of 1
19.What ids a dendrogram? - ANSWER✔️-A tree that represents clustering of
data
20.What is a paradigm? - ANSWER✔️-Belief framework that guides practice
in a certain field
21.What is a direct research tool? - ANSWER✔️-One involving interviews and
focus groups
22.What is an indirect research tool? - ANSWER✔️-One involving projective
techniques
and Computational Modelling Actual
Exam| 74 Real Questions With Detailed
Answers| Guaranteed Pass
1. What is the 10-way Cross Validation method used to do? - ANSWER✔️-
Measure models predictive accuracy and recognize how well a model will
predict data it has not seen.
2. How does the 10-way Cross Validation method work? - ANSWER✔️-
Divide data into 10 parts, train on 9, test on 1
3. What are decision trees split on? - ANSWER✔️-The best attribute to split
4. Does the accuracy of a decision tree increase as depth does? - ANSWER✔️-
Yes if same data is used for training and testing, no if cross validation test
used.
5. Can a decision tree handle both qualitative and quantitative data? -
ANSWER✔️-Yes
6. What are the three layers in an Artificial Neural Network? - ANSWER✔️-
Input Layer, Output Layer, Hidden Layer
7. What does Backpropagation in an ANN do? - ANSWER✔️-Begins with an
error output, propagates the error backward to calculate yield output error
values for neurons.
Then adjusts weights to better predict output from the same inputs.
Only applies to feed forward ANNs.
8. Values of x and y for a model can be referred to as? - ANSWER✔️-Input,
independent, dependent, output, explanatory, response
, 9. Can you run categorical variables through a linear model? - ANSWER✔️-
Yes, but can only give a binary outcome.
10. MLR with categorical variables yes (1) and no (0), what outcome gives yes?
- ANSWER✔️-Anything above 0
11.What transformations can be used on a linear regression model? -
ANSWER✔️-Log normal, cubic, square root, inverse squared
12.When fitting a model, R^2 should be... - ANSWER✔️-High
13.What does Backpropagation in an ANN NOT do? - ANSWER✔️-Makes the
hidden Layer visible.
14.What is the optimal number for hidden nodes for ANN? - ANSWER✔️-2-4
15.Why is scaling important? - ANSWER✔️-Pick clusters based on scale, so x
and y axis scales need to be constant in order to pick optimal cluster.
16.What is not true about scaling? - ANSWER✔️-Because scaling is in terms
of dimensions, it doesn't matter
17.Why is scaling done? - ANSWER✔️-To avoid bias
18.How are values normally scaled? - ANSWER✔️-To give a mean of 0 and
standard deviation of 1
19.What ids a dendrogram? - ANSWER✔️-A tree that represents clustering of
data
20.What is a paradigm? - ANSWER✔️-Belief framework that guides practice
in a certain field
21.What is a direct research tool? - ANSWER✔️-One involving interviews and
focus groups
22.What is an indirect research tool? - ANSWER✔️-One involving projective
techniques
