Module1: Introduction to Data Science

History of Data Science-Importance of Data Science in Modern Business-

Data Scientists-Data Science Activities in Three Dimensions-Overlapping
of Data Science with Other Fields- Data Analytic Thinking-Domains of
Application-Application of Computational Intelligence to Manage Data
Science Activities – Scenarios for Data Science in Business – Tools and
Techniques Helpful for Doing Data Science – Data Cleaning Tools – Data
Munging and Modelling Tools – Data Visualization Tools

Text Book:
Rajendra Akerkar and Priti Srinivas Sajja, Intelligent Techniques for Data
Science, Springer, 2016.
History of Data Science
Introduction
• Data Science is an interdisciplinary field that uses scientific methods,
processes, algorithms, and systems to extract insights from structured
and unstructured data.
• Its history spans mathematics, statistics, computer science, and
domain expertise.
Early Foundations (Pre-1900s)
• • 1600s: Probability theory by Blaise Pascal and Pierre de Fermat.
• • 1700s: Thomas Bayes formulates Bayes' Theorem.
• • 1800s: Development of statistical methods by Gauss and Laplace.
Statistical Era (1900–1960)
• • Advancements in statistical theory and sampling.
• • Introduction of correlation and regression analysis.
• • Use of punched cards for data processing.
Correlation-
Correlation is a statistical measure that describes the strength and
direction of a relationship between two variables.
Regression
Regression is a statistical method used to model and analyse the relationship
between a dependent variable (target) and one or more independent variables
(predictors).
Computer Age (1960–1990)
• • Rise of digital computers enabling large-scale data analysis.
• • Development of databases and SQL.
• • Machine learning concepts emerge.
Modern Data Science (1990–
Present)
• • Coining of 'Data Science' as a term in the 2000s.
• • Explosion of big data and cloud computing.
• • AI and deep learning revolution.
Timeline of Data Science
1600s 1700s 1800s 1900–1960 1960–1990 1990–Present
Probability Theory Bayes' Theorem Statistical Methods Statistical Era Computer Age Modern Data Science

Science-Importance of Data Science in Modern Business

Importance of Data
Science in Modern
Business
Driving Decisions, Innovation, and Growth
Introduction
• Data Science has become a crucial tool for modern businesses,
enabling data-driven decision-making, operational efficiency, and
customer personalization.
Key Benefits of Data Science in
Business
• • Informed Decision-Making
• • Predictive Analytics for Market Trends
• • Enhanced Customer Experience
• • Operational Efficiency
• • Competitive Advantage
Common Business Applications
• • Marketing Optimization
• • Sales Forecasting
• • Fraud Detection
• • Inventory Management
• • Product Recommendations
Data Science Workflow in Business
Data Collection Data Processing Analysis & Modeling Insights Generation Decision Implementation
Conclusion
• Data Science empowers modern businesses to stay competitive by
turning raw data into strategic insights. By leveraging data science,
organizations can make better decisions, innovate faster, and deliver
greater value to customers.
Data Science Activities
in Three Dimensions
Overlapping of Data Science with Other Fields
Introduction
• Data Science activities can be understood in three dimensions: data
analysis, data engineering, and domain expertise. It also overlaps with
related fields like statistics, computer science, and business
knowledge.
Three Dimensions of Data Science
• 1. Data Analysis – Extracting insights and patterns.
• 2. Data Engineering – Managing and processing large datasets.
• 3. Domain Expertise – Applying knowledge of the business or
application area.
Overlapping of Data Science with
Other Fields
Domain
Knowledge

Computer
Statistics
Science
Conclusion
• Data Science sits at the intersection of statistics, computer science,
and domain expertise. Success in the field requires a balance of all
three dimensions.
 Dimensions -
Overlapping of Data
Science with Other Fields
Exploring how Data Science intersects with various domains
Introduction
• Data Science is inherently multidisciplinary, drawing upon methods
and insights from multiple fields.
• Understanding the overlapping dimensions helps in leveraging the full
potential of Data Science.
• Key dimensions include: Statistics, Computer Science, Domain
Knowledge, and Communication Skills.
Overlapping Dimensions Diagram

Computer
Statistics
Science

Domain
Knowledge
Explanation of Overlaps
• Overlaps among these fields create powerful synergies:
• • Statistics + Computer Science → Machine Learning
• • Statistics + Domain Knowledge → Evidence-based Insights
• • Computer Science + Domain Knowledge → Automated Solutions
• • All three combined → Data Science Excellence
Data Analytic Thinking
- Domains of
Application
Application of Computational Intelligence to Manage Data Science
Activities
Data Analytic Thinking
• Data analytic thinking involves approaching problems with a mindset
geared towards extracting insights from data.
• Key aspects include:
• • Understanding the problem in a data-driven context
• • Identifying relevant data sources
• • Applying computational methods for analysis
• • Interpreting results for decision-making
Domains of Application
• Data Science applications span multiple domains:
• • Healthcare - Predictive diagnostics, patient monitoring
• • Finance - Fraud detection, risk modeling
• • Retail - Customer segmentation, demand forecasting
• • Manufacturing - Predictive maintenance, quality control
• • Transportation - Route optimization, traffic prediction
Application of Computational
Intelligence
• Computational Intelligence (CI) offers adaptive, learning-based
methods for managing data science activities.
• Examples include:
• • Machine Learning - Pattern recognition, prediction
• • Neural Networks - Deep learning for complex patterns
• • Evolutionary Algorithms - Optimization in dynamic environments
• • Fuzzy Systems - Handling uncertainty in data
Computational Intelligence in Data
Science Workflow
Data Collection Preprocessing ) Analysis

Interpretation &
Action
Data Analytic Thinking – Domains of
Application
Application of Computational
Intelligence to Manage Data Science
Activities
Introduction to Data Analytic
Thinking
• • Data Analytic Thinking involves problem-solving using data-
driven insights.
• • Focus on identifying patterns, relationships, and trends.
• • Helps in making informed decisions based on evidence.
Steps in Data Analytic Thinking
• 1. Define the problem clearly.
• 2. Identify and collect relevant data.
• 3. Preprocess and clean the data.
• 4. Apply analytical and computational models.
• 5. Interpret results and take action.
Domains of Application
• • Business Intelligence and Market Analysis
• • Healthcare Analytics
• • Financial Risk Management
• • Industrial Process Optimization
• • Education and E-learning Analytics
• • Government Policy and Decision Making
Computational Intelligence
Overview
• • A branch of AI that deals with adaptive mechanisms for
problem-solving.
• • Techniques include:
• - Neural Networks
• - Fuzzy Systems
• - Evolutionary Algorithms
• • Handles complex, uncertain, and imprecise data efficiently.
Application of Computational
Intelligence in Data Science
• Automating data preprocessing and feature selection.
• Optimizing predictive models for accuracy.
• Enhancing decision-making with adaptive algorithms.
• Real-time data analysis and anomaly detection.
Case Studies
• Healthcare: Predicting disease outbreaks using neural
networks.
• Finance: Fraud detection using fuzzy logic.
• Manufacturing: Quality control optimization with genetic
algorithms.
Conclusion
• Data Analytic Thinking bridges raw data and actionable
insights.
• Computational Intelligence boosts efficiency and
adaptability.
• Applicable across multiple industries for better decision-
making.
 Scenarios for Data Science in
Business – Tools and Techniques
Helpful for Doing Data Science
Prepared by: R Vignesh
Introduction
• Data Science in business transforms raw data into
actionable insights.
• Helps in improving decision-making, optimizing
operations, and predicting trends.
• Involves multiple tools and techniques for efficient
data handling.
Business Tool
Scenarios for Data Science in
Business
• Customer segmentation and targeting.
• Sales forecasting and demand prediction.
• Fraud detection and risk management.
• Product recommendation and personalization.
• Market basket analysis.
• Process automation and optimization.
Tools for Doing Data Science
• Programming Languages: Python, R, Julia.
• Data Visualization: Tableau, Power BI, Matplotlib,
Seaborn.
• Big Data Platforms: Hadoop, Spark.
• Databases: SQL, NoSQL.
• Machine Learning Frameworks: Scikit-learn,
TensorFlow, PyTorch.
Techniques for Doing Data Science
• Data Cleaning and Preprocessing.
• Feature Engineering.
• Predictive Modeling.
• Natural Language Processing.
• Time Series Analysis.
• Deep Learning and Neural Networks.
Case Studies
• E-commerce: Personalized product recommendations.
• Banking: Fraud detection using anomaly detection models.
• Retail: Demand forecasting for inventory optimization.
• Healthcare: Predictive analytics for patient readmission.
Conclusion
• Data Science empowers businesses to stay competitive.
• Tools and techniques should be chosen based on problem
requirements.
• Integration of analytics into business strategy ensures
sustainable growth.
 Data Cleaning Tools – Data
Munging and Modelling Tools –
Data Visualization Tools
Prepared by: R Vignesh
Data Cleaning Tools
• Open Refine – Data transformation and cleaning.
• Trifacta Wrangler – Interactive data cleaning.
• Talend – Data integration and cleaning.
• Data Ladder – Matching and deduplication.
Talend Data
Ladder

OpenRe
Trifacta
fine
Data Munging and Modelling Tools
• Python Pandas – Data wrangling and analysis.
• R – Data manipulation and modelling. R

• SAS – Advanced analytics and modelling.

• RapidMiner – End-to-end data science workflows.
RapidMi
ner

SAS

Pandas
Data Visualization Tools
• Tableau – Interactive dashboards.
• Power BI – Business intelligence reporting. Power
BI
• Matplotlib/Seaborn – Python visualization.
• [Link] – Web-based visualizations. Tableau

Matplot [Link]
lib
Conclusion
• Data Cleaning ensures quality input.
• Data Munging and Modelling enable accurate analysis.
• Visualization helps in interpreting and communicating results.

Mungin
Cleanin g
g

Visualiz
ation
Insights