Chapter 2 Quiz

Question 1

methods are trained on a set of training data and then their performance is evaluated on a separate set of validation data

Answer 1

data partitioning

Question 2

tasks of classification and prediction as well as pattern discovery

Answer 2

predictive analytics

Question 3

trying to predict value of categorical variable

Answer 3

classification

Question 4

trying to predict value of numerical variable

Answer 4

prediction

Question 5

finding general associations patterns between items in large databases through rules general to an entire population

Answer 5

association rules/affinity analysis

Question 6

method that uses individual users’ preferences and tastes given their historic purchases or measurable behavior indicative of preference

Answer 6

collaborative filtering

Question 7

consolidating a large number of records into a smaller set

Answer 7

data reduction

Question 8

methods for reducing the number of cases

Answer 8

clustering

Question 9

reduction of the number of variables

Answer 9

dimension reduction

Question 10

exploration by creating charts and dashboards

Answer 10

data visualization/visual analytics

Question 11

used in classification and prediction, must have data available in which the value of the outcome of interest is known

Answer 11

supervised learning algorithms

Question 12

data from which classification or prediction algorithm learns

Answer 12

training data

Question 13

sample of data where the outcome is known used for comparison between models

Answer 13

validation data

Question 14

sample of data where the outcome is known used to predict how well the model will do

Answer 14

test data

Question 15

there is no outcome variable to predict or classify

Answer 15

unsupervised learning algorithm

Question 16

steps in machine learning

Answer 16

understand project
obtain data
preprocess data
reduce dimensions (if necessary)
determine ML task
partition data
choose ML technique
perform task
interpret results
deploy model

Question 17

SEMMA

Answer 17

sample, explore, modify, model, assess

Question 18

character, integer, categorical

Answer 18

types of variables

Question 19

unordered categorical variables

Answer 19

nominal variables

Question 20

ordered categorical variables

Answer 20

ordinal variables

Question 21

categorical variables decomposed into a series of binary variables

Answer 21

dummy variables

Question 22

creating different binary dummy variables for more than one category

Answer 22

one-hot encoding

Question 23

values that lie far away from the bulk of the data

Answer 23

outliers

Question 24

knowledge of the particular application being considered

Answer 24

domain knowledge

Question 25

how to standardize a value

Answer 25

subtract mean from each value and divide by standard deviation

Question 26

how to normalize a value to [0,1]

Answer 26

subtract minimum value and divide by range

Question 27

creating a complex model that is too similar to the training data that it cannot be applied to other data

Answer 27

overfitting

Question 28

adds up squared errors so when an error is positive or negative it contributes the same

Answer 28

sum of squared errors

Question 29

average of the squared errors

Answer 29

mean squared error

Question 30

square root of the MSE to give idea of the typical error in the same scale used for the outcome variable

Answer 30

root mean squared error

Question 31

average of the absolute values of the errors

Answer 31

mean absolute deviation

Question 32

what do SSE, MSE, RMSE, and MAD measure?

Answer 32

prediction error