UBUS 288 FINAL EXAM

Intro to business data and analytics NIU comprehensive exam

Business Analytics

The process of analyzing data to gain valuable insights & inform business decisions

Descriptive Analytics

What has happened in the past?

Predictive Analytics

What could happen in the future?

Prescriptive Analytics

What should we do now?

For descriptive analytics we…

Gather, organize, visualize, and tabulate

For predictive analytics we…

Use statistical models to predict specific outcomes or likelihood

Data Privacy

Branch of data that is SECURITY related to the proper collection, usage, and transmission of data

What are the 3 key principles of data privacy?

Confidentiality, transparency, and accountability

Data Ethics

Studies moral problems related to data

AI

Aims to create machine capable of performing tasks requiring human intelligence

Generative AI

AI techniques that focuses on creating new content

Data

Compiled of facts, figures, and other content

Cross-sectional data

Record a characteristic of individual upon the same time

Time series data

Collected over several time periods focusing on certain groups

Structured/Tabular data

Has a predefined row-column format

Unstructured data

Do not conform to a predefined row-column format

Characteristics of Big Data

Volume, velocity, variety

Interval

0 is another point in the scale does not mean absence

Ratio

0 is a true point and represents complete absence

Nominal

No natural order

Ordinal

Contains a natural order

Discrete

Countable

Continuous

Measured, not counted

Data Wrangling

Transform raw data in a format that is easier to analyze

Data management

Process that is used to acquire, organize, store, manipulate, and distribute data

Mean

Sum of all observations divided by the size

Median

The middle value of sorted data

Mode

The most frequent observation

Percentile

Value by which the data falls under

5 number summary

Minimum, Q1:25th, Q2:50th (Median), Q3:75th, Maximum

Range

= Maximum-Minimum

Mean absolute difference

Calculates the absolute differences from the mean

Variance

Calculates the average of squared differences from mean

Standard Deviation

Square root of variance

Coefficient of variation

St. dev/ mean used to compare dispersion of data from many sets

Skewness coefficient

Measures asymmetry about the mean (focus on the tail)

Kurtosis Coefficient

Measures Tailness (short tail/long tail)

Excess Kurtosis

Compares tails from normal distribution

Covariance

Measures how two variables vary together in a direction

Correlation Coefficient

Measures the direction and strength of two variables

Sample Space

the set of all possible outcomes

Event

Subset of a sample space

Mutually exclusive events

Events that share no outcome

Mutually exclusive & exhaustive

Events that are mutually exclusive but together cover the entire sample space

Exhaustive event

The same as the sample space

The union of events

Contains outcomes in A OR B

Intersection of events

Outcomes must be both in A AND B

Complement of an event

Set of outcomes NOT in the event

What is the Range of probability ?

0<p<1

What are the 3 ways that probability is estimated ?

Subjective, Empirical, classical

Correlation Coeff. will always be between ?

-1 & 1

Complement Rule

P(A^c) = 1 - P(A)

The Additive Rule

P(A or B) = P(A) + P(B) - P(A and B)

Conditional Probability

the probability that one event happens given that another event is already known to have happened

Conditional Prob Rule

P(AIB)= P(A and B)/P(B)

Normal Distribution has a kurtosis of…

3

Linear Regression

Has a linear basis, draws a line that fits closest average distance to each point

Linear regression model

y = β0 + β1X

X:

Predictor, independent variable, feature, observation

Y:

Outcome, dependent variable, label

If the coefficient B1 is positive…

Relationship is + which means y is increasing

If the coefficient B1 is negative…

Relationship is - which means y is decreasing

If the coefficient B1 is zero…

There is no relationship Y is constant

Multiple Linear Regression

Incorporates more than one predictor and coefficient

Multiple Linear Regression model

y = β0 + β1X1 + β2X2 + ... + βnXn

Goodness of the fit

Evaluates the closeness between observed values and the values expected under a model

Coefficient of Determination also known as…

R²

R² definition

The proportion of variation in y that the model explains

If R² equals 1

Means the model explains all the variation in y perfectly

If R² equals 0

Means the model explains no variation in y (useless)

If P-value < Threshold

Significant

If P-value > Threshold

Not significant

P-value threshold is usually…

0.05

Interaction Equation

ŷ = b₀ + b₁X₁ + b₂X₂ + MX₁X₂

To model interaction…

We add a new term to the regression which is the product of two interacting variables

Interaction

Focuses on the combined effect of how the variables influences an outcome. Effect of 1 can depend on the level of another variable.

Dummy Variable

A numerical value used to represent categorical data, 0 or 1

Quadratic Equation

y= ax²+bx+c

Quadratic Phenomenon

When the direction of the effect changes from + to - or - to +

Quadratic Regression Model

y=β0​+β1​x+β2​x2+…..+ dx²1

If B2 < 0

Inverted U so a maximum

If B2 > 0

U-shaped so a minimum

To find the exact pt for optimization

X= - b1/ 2*b2

What is the main limitation of linear regression?

It assumes the relationships are linear

What is a measurement of dispersion?

IQR