1.
Introduction to Data Analytics
Data Overview
Data is raw facts (numbers, text, images). When processed, it becomes
useful information.
Types: Structured (Excel tables), Unstructured (social media posts), Semi-
structured (JSON/XML files).
State of the Practice in Analytics
Analytics is widely used in industries like healthcare (predicting
diseases), marketing (customer behavior), finance (fraud detection).
Growth drivers: AI, IoT, cloud computing, and cheaper storage.
Key Roles in Big Data Ecosystem
1. Data Engineer – Builds systems to collect and clean data (e.g., Hadoop
pipelines).
2. Data Scientist – Uses stats/ML to find insights (e.g., predictive models).
3. Business Analyst – Translates data insights into business decisions.
Memory Trick: "E.S.B" (Engineer → Scientist → Business Analyst).
2. Data Analytics Lifecycle
Phases (P-D-E-M-D-O)
1. Problem Identification – Define the business question (e.g., "Why are sales
dropping?").
2. Data Collection – Gather data from databases, surveys, sensors, etc.
3. Exploration & Cleaning – Handle missing data, remove duplicates.
4. Modeling – Apply algorithms (e.g., regression, clustering).
5. Deployment – Implement the solution (e.g., a recommendation engine).
6. Outcome – Review results and improve.
Global Innovation Network & Analysis (GINA)
A collaborative framework where teams share data/tools to solve big
problems (e.g., climate change analysis).
Exam Tip: Remember "People Don’t Eat Moldy Donuts, Okay?" (P-D-E-
M-D-O).
3. Data Mining Process
Definition
Extracting hidden patterns from large datasets (e.g., finding customer buying
habits).
Models
CRISP-DM (6 steps: Business Understanding → Deployment).
SEMMA (5 steps: Sample → Assess).
Steps (D-C-P-M-E-D)
1. Define Goal (e.g., "Reduce customer churn").
2. Collect Data (e.g., purchase history, complaints).
3. Preprocess (clean, normalize, remove outliers).
4. Model (use algorithms like decision trees).
5. Evaluate (check accuracy, e.g., 90% correct predictions).
6. Deploy (integrate into business systems).
Applications
Retail: Market basket analysis (e.g., "Customers who buy X also buy Y").
Banking: Fraud detection (e.g., unusual transactions).
Challenges
Privacy (GDPR compliance), noisy data, scalability.
Memory Trick: "Dirty Cats Prefer Milk Every Day" (Define, Collect,
Preprocess, Model, Evaluate, Deploy).
4. Basic Data Analytics Methods
Descriptive Analytics
"What happened?"
Tools: Dashboards, reports (e.g., monthly sales summary).
Diagnostic Analytics
"Why did it happen?"
Example: Drill-down analysis to find why sales dropped in Region X.
Predictive Analytics
"What will happen?"
Uses ML (e.g., predicting stock prices or customer churn).
Prescriptive Analytics
"What should we do?"
Example: Recommending the best marketing strategy.
Memory Trick: "Doctors Diagnose Patients Properly" (Descriptive →
Diagnostic → Predictive → Prescriptive).
5. Big Data & Business Intelligence Trends
What is Big Data?
Datasets too large/complex for traditional tools (e.g., social media data).
5V’s of Big Data
1. Volume (size in petabytes).
2. Velocity (speed of data generation, e.g., real-time tweets).
3. Variety (text, videos, logs).
4. Veracity (uncertainty/quality issues).
5. Value (extracting useful insights).
Goals of Big Data
Improve decisions (e.g., Netflix recommendations), reduce costs, innovate.
Big Data Architecture
Hadoop (stores data), Spark (fast processing), NoSQL (MongoDB for
unstructured data).
Future Trends
AI-driven analytics: AutoML (automated machine learning).
Edge computing: Processing data on devices (e.g., smart cameras).
Memory Trick: "Very Valuable Vegetables Vanish Quickly" (Volume,
Velocity, Variety, Veracity, Value).
Final Exam Strategy
✔ Definitions: Write 1-line explanations (e.g., "Predictive analytics forecasts
future trends").
✔ Examples: Relate to real-world use cases (e.g., "Amazon uses prescriptive
analytics for recommendations").
✔ Acronyms: Use memory tricks (P-D-E-M-D-O, 5V’s).
✔ Diagrams: Sketch the Data Analytics Lifecycle or CRISP-DM for extra
marks.
You’re all set! Revise this daily, and 80+ is guaranteed! 🎯
