Data Analytics Exam 2

Data exploration is critical for many purposes. Which one is NOT the right purpose?

look for univariate pattern, trend, and relationship

Data exploration is critical for many purposes. Which one is NOT the right purpose?

analyze only the imported data and no need to explore other data

Data exploration is critical for many purposes. Which one is NOT the right purpose?

Focus on the model fit and try to improve it

What can modern visualization solutions, such as SAS Visual Analytics and Tableau NOT do?

save time and efforts to analyze the data integrity and reliability

Data refinement is critical for many purposes. Which one is NOT the right purpose?

always include all columns into the refinement process

Bar charts are useful for categories or groups with high cardinality

False

Line charts are better than bar charts for comparing quantities across ordinal categories with low cardinality

False

In a scatter plot, values are plotted based on the Cartesian coor- dinates of ________ variables.

2

Scatter charts are an extension of bubble plots

False

Pie charts can be effective for illustrating a point when the number of slices is high (5–10)

False

Box plots group data by quartiles, where the top of the box corresponds to the 25th percentile (top quartile)

False

Tree maps display hierarchical data through the collection of nested rectangles.

True

A heat map is more useful for displaying data across multiple (hierarchical) categories.

False

A geo map uses the plot data of interest on a physical map.Not identifying the value

False

A correlation matrix uses shape to show how strong the linear relationship is between different measure variables.

False

Clustering is a machine learning approach for grouping data with dissimilar underlying characteristics.

False

Clustering is used to explore data with requiring a specific outcome variable, that is, it is supervised learning.

False

There are two most common methods for clustering data. Which one is NOT?

n-means clustering

Machine learning enables many business applications. Which one is NOT?

medical treatment

Which one is NOT a canonical method of supervised learning?

Clustering

What machine learning method is often used to classify if email is spam or whether a customer churned?

classification

___________ is used to mathematically map a set of input data to an output that is continuous or to a discrete output

Regression

_____________ helps firms to identify meaningful customer segments, allowing them to target defined group

Clustering

Firms need to utilize segments in predictive modelling by creating a same predictive model for all segments

False

Segmentation is now being used for a variety of analytic tasks. Which one is NOT?

transportation

Similar to recommendations, ______ is essential for targeted advertisements.

Clustering

Clustering algorithms take raw data and produce clusters using measures of ______ to quantify similarity in the data.

Distance

Clustering uses two most common methods. Which one is NOT?

horizontal clustering

Hierarchical clustering uses a _____ process to develop clusters

Bottom-up

Hierarchic clustering is the most common method for determining clusters, due to its scalability to large datasets.

False

__________ measures distance using standard deviations.

Mahalanobis distance

The cluster matrix is a matrix of different variables, and show the association between these variables

True

Which European country has cultural dimensions that are closer to Asian countries, such as China and India?

France

Clustering assumes that people fit into natural groupings and reflect people’s

characteristics in reality.

False

________of the detailed results provides a summary of the model and an F-test statistic

The Overall ANOVA

When conducting parameter estimates, a ______ test is used to test whether a predictor is significant.

t-test

To create a regression equation we need to look at the ___________.

Perameter estimates

In reporting p-values we follow the standard convention that a p-value of ______ is flagged

<.05

The most common way of assessing the fit of a model is the _______ statistic

R-squared

The _______is the amount of variation in the dependent variable that is captured by the independent variables

R-squared value

A model with 6 predictors and a sample size of 30, the adjusted R-square will be _____ than the unadjusted value.

Lower

SAS VA provides further fit statistics such as the ________

AIC

To dig deeper into linear regression, which one do you NOT need to learn?

Degrees of accuracy

The null hypothesis proposes that there is a __________ association between two

variables

No

Can we reject the null hypothesis at some level of significance? By convention, we use _____ as the cut-off.

0.05

What type of error occurs when we fail to reject a false null hypothesis?

type 2

The systematic variation is divided by the unsystematic variation to produce the ______

test statistic

What is the degree of freedom?

sample size -1

______________ is the ratio of the variance explained by the model to the total variance to be explained

R-square

The ________answers: does the model perform better than simply using the mean value as the estimate?

F statistic

With _______ we add more predictors and the regression formula consequently has more regression coefficients

multiple regression

The multiple regression assumes that __________________ are normally distributed

residuals

To help inspect the distribution of the residuals SAS VA produces a______ chart of studentized (standardized) residuals

scatter plot

Multicollinearity can be assessed using the ________________

VIF

Which is NOT the commonly used transformation for non-normalized residuals?

Squared roots of data transformation

_________ calculates test statistic of interest by taking many samples of the dataset, calculating for each sample

Bootstrapping

Given a normal distribution we would expect that 99.9% should lie within _____

+/- 3.29

Cook’s distance values of greater than ________ are generally accepted to indicate an influential observation.

0.3

In linear regression, ____________ is an observation with a large residual (error term).

outlier

A box and whisker plot for a dataset will MOST clearly show:

If the data is skewed and, if so, in which direction.

Which of the following BEST describes the data and information flow within an organization?

Information architecture

A preferred method or best practice for organizing data in a data warehouse for reporting and analysis is:

Multidimensional modeling

When analyzing responses of a survey of why people like a certain restaurant, factor analysis could reduce the dimension in which of the following ways?

Collapse several survey questions regarding food taste, health value, ingredients, and consistency into one general unobserved “food quality” variable.

A clothing company wants to use analytics to decide which customers to send a promotional catalog in order to attain a targeted response rate. Which of the following techniques would be the MOST appropriate to use for making this decision?

Logistic regression

Which of the following is an effective optimization method?

Mixed-integer programming (MIP)

A furniture maker would like to determine the most profitable mix of items to produce. There are well-known budgetary constraints. Each piece of furniture is made of a predetermined amount of material with known costs, and demand is known. Which of the following analytical techniques is the MOST appropriate to solve this problem?

Optimization

A company ships products from a single dock at their warehouse. The time to load shipments depends on the experience of the crew, products being shipped, and weather. The company thinks there is significant unmet demand for their products and would like to build another dock in order to meet this demand. They ask you to build a model and determine if the revenue from the additional products sold will cover the cost of the second dock within two years of it becoming operational. Which of the following is the MOST appropriate modeling approach?

Discrete-event simulation, because there is a sequence of discrete random events through time.