visual analytics exam 3

What is SQL?

Structured query language

What is a "relational database"?

What enterprise systems run on top of

• The tables can be connected to other tables based on common columns

One-to-one

a link between the information in two tables, where each record in each table only appears once.

One-to-many

 one record in a table can be associated with one or more records in another table

Many-to-many

when one or more items in one table can have a relationship to one or more items in another table when one or more items in one table can have a relationship to one or more items in another table

What is data mining?

the process of sorting through large data sets to identify patterns and relationships that can help solve business problems through data analysis

What is unsupervised learning?

consist of descriptive data-mining methods in which there is no outcome variable to predict.

Market segmentation

is a process commonly used to divide customers into different homogenous groups

k-means clustering

iteratively assigns each observation to one of k clusters composed of similar observations.

Hierarchical clustering

starts with single-observation clusters and sequentially merges the most similar ones to create a set of nested clusters.

What is text mining?

the process of extracting information from text data

Document

 a contiguous piece of text

Terms

words

Corpus

(the body of text) a collection of text documents to be analyzed

bag of words

treats every document as a collection of individual words or terms

presence/absence document-term matrix

 aka binary document term matrix, rows represent documents and column represent words

Tokenization

is the process of dividing text into separate terms

term normalization

 the process of identifying tokens

Stopwords

 “the”, “and”, “of” are removed from the document and all letters are converted to lowercase

Stemming

 the process of converting different forms of the same word and synonyms to its stem or root word

frequency term-document matrix

• The rows represent the documents 

• The columns represent the tokens

• The entries in the matrix are the frequency of occurrence of each token in each document

Word cloud

• is a visual display that contains the key terms of a document

What is regression?

• is a set of data mining methods having the task of predicting a quantitative outcome using a set of input variables

Features

•  task of predicting a quantitative outcome using a set of input variables

What is supervised learning?

Predictive data mining methods, use known values of the outcome variable to “supervise” the learning process of how to use the input features to predict future values of the outcome variable

• Data sampling, data preparation, data partitioning, model construction and assessment

What is a neural network?

•  a predictive data mining method whose structure is motivated by the biological functioning of the brain

What is classification?

• is a set of data mining methods having the task of predicting categorical outcome variable using a set of input variables, or features.

What is a confusion matrix?

• a cross-tabulation of the actual class of each observation that displays a model’s correct and incorrect classifications. 

Class 1 Error

False negative

Class 0 Error

False positive

What is a classification tree?

•  is a predictive model consisting of a sequence of rules on the input features.

What is analytics?

• Data, Information, Knowledge, Action (DIKA)

Production

 assembly/manufacturing

Procurement

ordering from a supplier/vendor

Fulfillment

delivering the product/having and fulfilling customer order

Descriptive Analytics

describing from the past (scorecard)

Predictive Analytics

forecast/weather - what will happen

Prescriptive Analytics

• what should you do- plan of action, optimization model

Digitization

paper to computer/ digital format

Digitalization

continuously improving digital form

4 Vs of Big Data

• volume, variety, velocity, veracity

Decision-making process

1. Identify and define the problem

2. Determine the criteria that will be used to evaluate alternative solutions.

3. Determine the set of alternative solutions.

4. Evaluate the alternatives.

5. Choose an alternative.

Data Flow Diagram

 a process model used to depict the flow of data through a system and the work or processing performed by the system.

DFD Symbols

Data flow

1 piece of data

Composite data flow

multi data pieces together in one data flow

ex. receipt

Context Diagram (level 0 diagram)

Decomposition Diagram

High-level Diagram (level 1 diagram)

Assembling

Components - nuts, bolts, wheels-> skateboard

Manufacturing

Raw materials- plastic pellets -> plastic plate

Discrete

distinct items (countable). Usually identifiable (bill of materials)

Process Manufacturing

like a recipe, can not be easily disassembled

Instance Level Information

status of production for a particular order, required inventory status report, stock requirements list (at that moment) (status)

 Process Level Information

average time to produce, how many on time/delayed, reasons for delay 

Intra-

Inside/within (Human Capital Management, Asset Management, Human resources)

Inter

 between, among companies (Supply Chain Management, Supplier relationship management)

Preattentive attributes

features that can be used in a data visualization to reduce the cognitive load required to interpret it

• Include color, shape, size, length

Data-ink

 the ink used in a table or chart that is necessary to convey the meaning of the data to the audience

Data-ink Ratio

the proportion of ink used for data to the total amount of ink in a table or chart (high ratio is good)

Table Design Principles

• Keep the data-ink ratio high

• Use lines only to separate labels from data and calculated fields

• Labels should be left-aligned

• Values should be right-aligned

• Center vertical labels

What is a data dashboard?

• a data visualization tool that illustrates multiple KPIs and automatically updates as new data becomes available

Objective of data wrangling

•  to produce a final dataset that is accurate, reliable, and accessible

Activities of data wrangling

• Cleaning, managing, transforming

• Right data, right place, right time, right form

Structured Data

• refers to data arrayed in a predetermined pattern to make them easy to manage and search (flat file is most common pattern)

Unstructured Data

data not arranged in a predetermined pattern

Semi-structured Data

• not organized as structured data but contain elements that allow for isolating some raw data elements

How do we assess risk?

Likelihood (probability)

Impact (severity)

Likelihood

probability

Impact

severity

The only document that contains weight is

packing list

In Tableau, tables are connected

by common fields

rows

records, observations

fields

columns

Query aggregation operators

In sql, a group by clause always includes 

-Count, avg, min, max, or sum

Logical operators

more than one condition

• and, or, not, between, exists, or in 

Small and Medium Manufactures

make up 95%

CRM

Customer relationship management

PO

purchase order

SOM

serviceable obtainable market

TDP

total distribution points

SQL

Structured Query Language

DFD

data flow diagram

PPF

production possibilities frontier

DIKA

Data, Information, Knowledge, Action