🧠 Descriptive Statistics I

📊 Key Concepts

 Data: Facts and figures for analysis.

 Variable: A measurable characteristic.
 Observation: A set of values for variables.
 Random Variable: Outcome not known with certainty.
📈 Types of Data

 Quantitative: Numerical (e.g., income).

 Categorical: Non-numeric (e.g., gender).
 Cross-sectional: One time point.
 Time-series: Over time.
📋 Data Collection

 Experimental: Variables manipulated.

 Observational: No manipulation (e.g., surveys).
📍 Measures of Central Tendency

 Mean: x=
∑ xi
n
x n +x
 Median: Middle value. Median= ( 2 ) ( n2 +1 ) (even sample size)/ Median=x (n +1 )
2
2
 Mode: Most frequent value.
📊 Data Analysis in Excel

 Sorting & Filtering

 Conditional Formatting
 =AVERAGE(), =MEDIAN(), =MODE.MULT()
📐 Frequency Distributions
frequancy
 Relative Frequency:
n
 Percent Frequency: Relative × 100
 Histogram: No gaps, shows frequency.
 Frequency Polygon: Line graph version of histogram.
🧮 Descriptive Statistics II

📉 Cumulative Distributions

 Cumulative Frequency: Add up frequencies.

Cumulative Freq
 Relative:
Total

 Percent: Relative × 100

📊 Skewness

 Right-Skewed: Tail on the right.

 Left-Skewed: Tail on the left.
 Symmetrical: No skew.
Geometric mean
Let x 1 , x 2 ,....... , x n be a set of n positive values .
1
o Formula: x g =( x 1 ⋅ x 2 ⋅ …⋅ x n ) n

(∏ )
n 1
n
o Alternate form using product notation: x g = xi
i=1

o Percentage Growth Rate (e.g. for investment):

( )
1
Final Value n
Annual Growth Rate= −1
Initial Value

📉 Measures of Variability

 Range: Max - Min

 Variance:

o Sample: s =
∑ (
2
x i−x )2
n−1

o Population: σ 2=
∑ ( xi −μ )2
N

 Standard Deviation: s= √ s2

 Coefficient of Variation (CV):

s
CV= ×100 %
x

📊 Position Measures
 p
 Percentile (Lp): L p= ( n+ 1 )
100
 Quartiles:
o Q1 = 25th percentile

o Q2 = Median

o Q3 = 75th percentile

o IQR = Q3 - Q1

📏 z-Score
x−x
 z=
s
 Measures how many standard deviations a value is from the mean.
📐 Empirical Rule

 For bell-shaped distributions:

o 68% within ±1σ

o 95% within ±2σ

o 99.7% within ±3σ

🚨 Outliers

 Typically: ∣z∣>3|z| > 3

📦 Boxplots

 Visual summary using min, Q1, median, Q3, and max.

 Useful for comparing datasets.
📉 Scatter Plot

 Visualizes relationship between two variables.

🔄 Covariance & Correlation

 Covariance: s xy =
∑ ( x i−x ) ( y i− y )
n−1
s xy
 Correlation: r xy =
sx s y

σ xy
 ρ xy =
σxσ y
 o Range: -1 to +1

🤖 AI in Business Analytics

🤖 What is AI?

 Simulates human intelligence (learning, planning, reasoning).

 Applied in business for automation, prediction, and insights.
📈 How AI Helps Business Analytics

 Automates repetitive tasks.

 Enhances decision-making with predictive analytics.
 Customizes marketing and user experience.
 Processes unstructured data via NLP.
 Tracks risks, monitors progress in real-time.
⚙️Key Components of AI in Analytics

 Machine Learning (ML)

 Deep Learning
 Natural Language Processing (NLP)
 Predictive Analytics
 Model Evaluation (e.g., F1-score, precision)
 Ethical AI: Focuses on fairness, transparency, accountability.
🔮 Future Trends

 Augmented Analytics
 AutoML
 AI-Driven Development
 Edge AI: AI on local devices.
 Ethical AI: Responsible, bias-aware AI.
⚠️Challenges

 Skill gaps
 Integration with existing systems
 Ethical concerns
💼 Role of Business Analysts in AI Era

 Focus on strategic, creative tasks

 Upskill in AI/ML
 Collaborate with data scientists

📘 INTRODUCTION TO DATA ANALYTICS

1. Definition of Data Analytics

 Process of collecting, transforming, and interpreting data to support decision-making.

 Involves extracting insights, predicting outcomes, and forming strategies.

2. Key Terms Comparison

Term Focus/Scope
Data Analytics Broad term; includes data management, analysis, tools.
Data Analysis Subset of analytics; focuses on cleaning, modeling, querying.
Includes analytics + ML, AI; predicts outcomes and builds
Data Science
models.
Business Analytics Uses data insights to make business decisions; process-driven.
Business Intelligence (BI) Descriptive, focuses on historical data and KPIs.

3. Types of Data Analytics

Type Description Question Answered

Descriptive Describes past data What happened?
Predictive Uses past data to forecast What will happen & why?
Prescriptive Suggests actions based on predictions How to make it happen?

4. Analytics Tools

 Data mining, visualization, reporting, modeling, dashboards.

 Common software: Excel, Power BI, Tableau, R, Python.

5. Big Data

 Defined by 4Vs: Volume, Velocity, Variety, Veracity.

 Requires tools like Hadoop, MapReduce.
 Enables AI applications, e.g. facial recognition, self-driving cars.

6. Decision-Making Levels

Level Focus Responsibility

Strategic Long-term, overall goals Top management
 Level Focus Responsibility
Tactical Implementation strategies Mid-level managers
Operational Daily operations Operations/frontline managers

7. Applications Across Sectors

 Finance: Portfolio optimization, risk forecasting.

 HR: Talent retention, performance prediction.
 Marketing: Customer insights, ad targeting.
 Healthcare: Scheduling, diagnosis, inventory.
 Web: Website configuration, A/B testing.
 Supply Chain: Routing, risk analysis.
 Sports: Player selection, ticket pricing.
 Gov & NGOs: Tax fraud detection, resource allocation.

8. Ethical & Legal Issues

 Ethical: Transparency, fairness, privacy protection.

 Legal: GDPR – consent, usage clarity, right to data.
 INFORMS Guidelines: Accountability, honesty, respect.

📘 DATA ANALYTICS LIFE CYCLE

1. Definition

 A structured process for managing a data analytics project from start to

implementation.

2. Importance

 Helps align data efforts with business goals.

 Provides clarity, reduces errors, enhances decision-making.

3. Phases of the Life Cycle

🔍 1. Discovery

 Understand business problem & domain.

 Identify stakeholders and success criteria.
 Develop hypotheses and identify data sources.

🧹 2. Data Preparation

 Clean and transform data (remove duplicates, fix errors).

 Create analytic sandbox (test environment).
 Tools: OpenRefine, Hadoop, Excel, R.
🧠 3. Model Planning

 Choose analytical techniques (e.g., regression, classification).

 Explore data relationships.
 Techniques: ETL, ELT, R, SQL, SAS.

4. Model Building

 Train and test models.

 Validate model performance.
 Tools: Power BI, Tableau, RapidMiner, Python, MATLAB.

5. Communicate Results

 Compare model outcomes with success criteria.

 Prepare reports for stakeholders.
 Emphasize key findings and business impact.

🚀 6. Operationalization

 Deploy model (pilot → full scale).

 Monitor performance in real environment.
 Final report + documentation submitted.

4. Key Roles in Analytics Project

Role Responsibility
Business User Provides domain context and feedback
Project Sponsor Funds project, defines scope and value
Project Manager Tracks progress, deadlines, quality
BI Analyst Builds dashboards, understands KPIs
Database Admin (DBA) Manages access, ensures data security
Data Engineer Manages pipelines, prepares data
Data Scientist Builds models, interprets data

5. Benefits of Structured Lifecycle

 Better risk management

 Faster, optimized decision-making
 Improved efficiency and business alignment