SDDS701PC: Predictive Analytics
UNIT – I
Introduction to Predictive Analytics
1. Introduction to Analytics
Analytics is the process of using computational methods to discover and report influential
patterns in data. The goal of analytics is to gain insight and often to affect decisions. Data is
necessarily a measure of historic information so, by definition, analytics examines historic data.
Analytics is the process of collecting, organizing, analyzing, and interpreting data to make better
decisions. It helps organizations understand past events, monitor current performance, and
predict future outcomes.
Real-World Example
A supermarket analyzes customer purchase data to understand which products sell most during
festivals. This helps them stock items efficiently and increase profits.
Predictive Analytics
Predictive Analytics is the branch of data analytics that uses historical data, statistical
techniques, data mining, and machine learning algorithms to predict future events or outcomes.
Predictive analytics is the process of discovering interesting and meaningful patterns in data. It
draws from several related disciplines, some of which have been used to discover patterns in data
for more than 100 years, including pattern recognition, statistics, machine learning, artificial
intelligence, and data mining.
Example 1: Online Shopping
When you shop on Amazon, it recommends products based on your previous purchases and
browsing history.
Prediction: Which product you are likely to buy next.
Example 2: Banking
Banks analyze customer financial history to predict whether a customer will repay a loan.
Prediction: Loan approval or rejection.
Example 3: Healthcare
Hospitals use patient records to predict the risk of diseases.
Prediction: Chances of developing diabetes or heart disease.
Example 4: Education
Colleges analyze attendance, assignments, and internal marks.
Prediction: Whether a student will pass or fail the examination.
Types of Analytics
Analytics is the process of examining data to gain insights and support decision-making.
Analytics is generally classified into four types:
1. Descriptive Analytics
2. Diagnostic Analytics
3. Predictive Analytics
4. Prescriptive Analytics
These types represent the evolution from understanding the past to making intelligent decisions
for the future.
1. Descriptive Analytics
Definition
Descriptive Analytics summarizes historical data to understand what has happened in the past. It
converts raw data into meaningful information through reports, charts, dashboards, and
summaries.
Explanation
Organizations collect large amounts of data every day. Descriptive analytics analyzes this data
and presents it in a simple form for understanding trends and performance.
Example
A college analyzes semester examination results and finds that 85% of students passed and 15%
failed.
A company reviews last year's sales report and observes that sales increased by 20%.
Techniques Used
Reports
Dashboards
Data Visualization
Data Aggregation
Advantages
Easy to understand.
Provides a clear view of past performance.
Supports business reporting.
Helps identify trends and patterns.
Disadvantages
Only explains past events.
Cannot identify causes.
Does not predict future outcomes.
Important Points
Most common form of analytics.
Uses historical data.
Foundation for all other analytics.
Answers "What happened?"
2. Diagnostic Analytics
Definition
Diagnostic Analytics investigates data to determine why an event occurred. It identifies the root
causes of problems and opportunities.
Explanation
After identifying a problem through descriptive analytics, diagnostic analytics helps discover the
reasons behind it.
Example
A company notices a decline in sales during a particular month.
Diagnostic analysis reveals:
Increased competition.
Poor marketing campaign.
Seasonal demand reduction.
Techniques Used
Drill-down Analysis
Correlation Analysis
Root Cause Analysis
Data Discovery
Advantages
Identifies causes of problems.
Supports informed decision-making.
Helps improve business processes.
Disadvantages
Requires detailed data.
Analysis may be complex.
Time-consuming.
Important Points
Focuses on cause-and-effect relationships.
Helps solve business problems.
Answers "Why did it happen?"
3. Predictive Analytics
Definition
Predictive Analytics uses historical data, statistical methods, and machine learning algorithms to
predict future outcomes.
Explanation
Predictive analytics identifies patterns in historical data and uses them to forecast future events.
Example
A bank predicts whether a customer will repay a loan based on:
Income
Credit Score
Employment History
An e-commerce website predicts products customers are likely to purchase.
Techniques Used
Machine Learning
Regression Analysis
Classification Algorithms
Neural Networks
Decision Trees
Advantages
Helps anticipate future events.
Improves decision-making.
Reduces risks.
Increases profitability.
Disadvantages
Requires high-quality data.
Predictions may not always be accurate.
Complex models require expertise.
Important Points
Uses historical data for forecasting.
Widely used in business and healthcare.
Answers "What will happen?"
4. Prescriptive Analytics
Definition
Prescriptive Analytics recommends actions that should be taken based on predictions and
business objectives.
Example
Google Maps predicts traffic conditions and suggests the fastest route.
Airlines adjust ticket prices based on demand predictions.
Techniques Used
Optimization Algorithms
Simulation Models
Artificial Intelligence
Machine Learning
Advantages
Provides actionable recommendations.
Improves efficiency.
Optimizes resources.
Supports strategic planning.
Disadvantages
Most complex type of analytics.
Requires advanced technology.
High implementation cost.
Key Applications of Predictive Analytics
Across various industries, predictive analytics is applied to solve complex problems and drive
smarter decision-making:
Finance: Credit risk assessment and fraud detection by identifying anomalous, suspicious
transactions.
Healthcare: Forecasting disease outbreaks, predicting patient hospital admissions, and
identifying individuals at risk of specific conditions.
Retail: Optimizing supply chains, predicting seasonal sales, managing inventory, and
personalizing customer recommendations.
Logistics: Route optimization to save time and energy, and anticipating shipping delays
Applications include:
Fraud Detection: Banks and e-commerce platforms analyze transaction anomalies in real-time
to flag irregular patterns before theft occurs.
Customer Relationship Management (CRM): Retailers predict buying behavior and customer
churn to deliver personalized product recommendations and targeted retention campaigns.
Preventive Manufacturing: Industrial tools forecast equipment failure before it happens,
minimizing costly downtime and production delays.
Healthcare Diagnosis & Care: Providers analyze patient records to identify individuals at high
risk for readmission or chronic illnesses, allowing for early intervention.
Supply Chain & Inventory: Businesses optimize stock levels and forecast disruptions based on
historical performance and demand cycles.
Risk & Credit Scoring: Financial institutions measure credit risk by predicting the likelihood of
loan defaults based on similar historical borrower profiles.