1 UNIT 1 INTRODUCTION TO DATA ANALYTICS

Lesson 1: learning, and data visualiza�on to derive

What is Data Analytics? ac�onable insights and support decision- 2. Data Coection
making. Purpose:
Data B. Scope: Broader in scope, encompassing • Gather data relevant to the defined
raw, unorganized that need to be processed descrip�ve, diagnos�c, predic�ve, and objec�ves, iden�fied as data
prescrip�ve analy�cs. Answers ques�ons requirements.
Information like "What will happen?" and "What Key Questions :
data is processed, organized, structured or should we do?" using advanced - What data is needed?
presented in a given context methods. - Where can the data be sourced?
• Analyzing average sales trends to - How will the data be collected?
predict future demand. - Is the data reliable and relevant?
C. Goals: To generate insights and forecasts
Data Analysis that can guide future strategies and long- 3. Data Preparation/ Cleaning
A process that begins with the collec�on of term decision-making. Purpose :
raw data. This data is then organized and D. Applica�on: To generate insights and • Clean and organize data for analysis.
summarized to extract meaningful forecasts that can guide future strategies • Preparing data for analysis by
informa�on. Through detailed analysis, and long-term decision-making. removing or modifying data that is
paterns and insights are derived, which are • Analyzing survey responses to incorrect, incomplete, irrelevant,
synthesized into ac�onable conclusions that identify customer preferences duplicated, or improperly formated.
drive decisions. The progression emphasizes E. Techniques: Advanced statistical Key Questions :
how data transi�ons from raw form to methods, machine learning, data mining, - What data needs to be cleaned or
meaningful insights, ul�mately enabling predictive modeling, and optimization. transformed?
informed ac�ons. F. Tools: Python, R, Tableau, Power BI, SQL, - How will missing or invalid data be
Hadoop, Spark, and cloud platforms like handled?
Data Analysis vs Data AWS or Google Cloud. - How will data be standardized or
Analytics formated?
Data Analysis Data Analysis vs Data Analytics - Are there any outliers or errors in the
A. Defini�on: The process of collec�ng, Data Analysis is a subset of Data Analy�cs. data?
organizing, and interpre�ng raw data to While data analysis (Insight) focuses on
iden�fy paterns, rela�onships, and examining and interpre�ng exis�ng data to 4. Data Analysis
trends. It focuses on extrac�ng uncover paterns, trends, and insights about Purpose :
meaningful insights from data for what happened, data analy�cs (Strategy) • Analyze data to uncover paterns,
specific purposes. takes a broader approach by using advanced trends, and insights, and build models
B. Scope: More focused and specific. tools and methods to process data, uncover to address the objec�ves.
Typically deals with historical data to trends, and predict future outcomes. Key Activities/Processes:
answer ques�ons like "What happened?" Encompasses the en�re process, including ✓ Explora�on
planning, analyzing, and predic�ng
or "Why did it happen?” ✓ Visualiza�on
• Calculating average sales outcomes, to guide future ac�ons and
✓ Modeling
C. Goals: To understand and summarize the decision-making.
Key Questions:
current or past state of data, helping to - What paterns or trends are present
make immediate or retrospec�ve Lesson 2:
in the data?
decisions. The Analytics Process - How can the data be visualized to
D. Applica�on: Used in research, academic highlight insights?
studies, and one-�me repor�ng tasks. 1. Problem Definition
- What type of model or analysis is
• Summarizing survey responses for a Purpose :
most appropriate?
market research report. • Clearly define the problem or
- How can we validate and evaluate the
E. Techniques: Descriptive statistics, ques�on you aim to solve and
findings?
correlation analysis, hypothesis testing. establish the scope of the analysis.
- What insights can be derived from the
F. Tools: Excel, SPSS, R (basic statistical • This sets the founda�on for the en�re
analysis?
functions), Python (e.g.,Pandas). analy�cs process.
Key Questions:
5. Interpretation of Results
Data Analytics - What problem are you solving?
Purpose:
A. Defini�on: A broader field that - What decisions will the insights
• Translate analysis into ac�onable
encompasses data analysis but also inform?
insights.
includes the use of tools, technologies, - What to measure? How to measure
• Relate the findings to the original
and methodologies to collect, clean, it?
problem and provide ac�onable
process, and analyze data. It o�en
recommenda�ons.
involves predic�ve modeling, machine
Key Questions:
2 UNIT 1 INTRODUCTION TO DATA ANALYTICS

- What do the results mean in the 4. Prescriptive Analysis

context of the problem? • How can we make it happen?
- Are the results consistent with • Purpose: Recommends ac�ons to
expecta�ons or exis�ng knowledge? op�mize outcomes based on
- What are the implica�ons of the predic�ons and scenarios.
results? • Example: Op�mizing delivery routes
- What are the limita�ons or to reduce shipping costs.
uncertain�es in the results? • Common Techniques:
- How can the results be communicated - Op�miza�on algorithms
effec�vely? - Decision trees
- Simula�on models.
6. Implementation and Iteration
Purpose: Lesson 4:
• Relate the findings to the original Importance/
problem and provide ac�onable
recommenda�ons. Lesson 3:
Applications of
Key Questions:
Types of Analytics Analytics
- How will the findings be applied to
address the problem? Importance of Analytics
1. Descriptive Analysis
- What resources are needed for • Informed Decision-Making:
• What happened?
implementa�on? Analy�cs provides evidence-based
• Purpose: Summarizes and interprets insights that allow organiza�ons to make
- How will the implementa�on process
historical data to iden�fy paterns or more precise decisions, reducing reliance
be monitored and measured?
trends. on intui�on.
- What feedback or results from the
• Example: Tracking the average daily
implementa�on are needed for • Problem Solving:
number of customers visi�ng a store
itera�on? It iden�fies trends, paterns, and
over the past year. anomalies, enabling quicker and more
Life Cycle of Analytics • Common Techniques: effec�ve responses to problems.
- Data aggrega�on/ Data
• Eiciency and Optimization:
Summariza�on
By analyzing opera�ons and workflows,
- Data visualiza�on
analy�cs helps op�mize resource
alloca�on and improve produc�vity.
2. Diagnostic Analysis
• Competitive Advantage:
• Why did this happen? Organiza�ons using analy�cs effec�vely
• Purpose: Examines historical data to can predict market trends, understand
iden�fy the root causes behind a customer needs, and stay ahead of
par�cular outcome. compe�tors.
• Example: Inves�ga�ng root causes of • Risk Management:
increased complaints in customer It an�cipates poten�al risks and develops
support. strategies to mi�gate them.
• Common Techniques:
- Drill-down analysis/ Data Importance of Analytics in
Decomposi�on Specific Fields
- Sta�s�cal analysis (Correla�on and A. Education
regression) • Improving Teaching Strategies:
Analy�cs evaluates teaching methods'
3. Predictive Analysis effec�veness, iden�fying which
• What will happen? approaches yield beter student
• Purpose: Uses historical data to outcomes.
predict future outcomes or trends. • Student Performance Tracking: Helps
• Example: Forecas�ng sales for the iden�fy struggling students early and
next quarter. provides tailored interven�ons.
• Common Techniques: • Curriculum Development: Guides the
- Machine learning models design of courses and learning
- Timeseries analysis materials based on data-driven insights
- Predic�ve modeling into student needs.
3 UNIT 1 INTRODUCTION TO DATA ANALYTICS

B. Healthcare  Pros: Strong in sta�s�cal compu�ng, 3. Databases: SQL for querying (MySQL,
• Pa�ent Care: Predic�ve analy�cs open-source PostgreSQL).
iden�fies high-risk pa�ents and  Cons: Steeper learning curve compared
personalizes treatment plans. to Python II. Diagnostic Analytics
• Opera�onal Efficiency: Helps manage 4) Microso� Power BI 1. Sta�s�cal So�ware: Excel Analysis
resources like staff, equipment, and  Type: Business analy�cs service ToolPak, SPSS.
beds effec�vely.  Usage: Data visualiza�on, repor�ng 2. Programming: Python (pandas,
• Public Health: Tracks disease outbreaks  Pros: Integrates with other Microso� matplotlib), R (ggplot2).
and evaluates interven�ons. products, user-friendly 3. Explora�on Pla�orms: Tableau,
 Cons: Limited advanced analy�cs Power BI (with drill-down
C. Busine and Marketing capabili�es capabili�es).
• Customer Insights: Analyzes purchasing 5) Tableau
behavior to develop targeted marke�ng  Type: Data visualiza�on tool III. Predictive Analytics
campaigns.  Usage: Interac�ve dashboards, data 1. Spreadsheet Models: Excel (basic
• Inventory Management: Predicts storytelling forecas�ng with built-in tools).
demand paterns to reduce wastage  Pros: Intui�ve interface, powerful 2. Machine Learning Libraries: Scikit-
and meet consumer needs. visualiza�on capabili�es learn (Python), caret (R).
• Strategic Planning: Guides long-term  Cons: Can be expensive, requires 3. Cloud Services: Google AutoML, AWS
strategies through market trend training Forecast (easy-to-use pla�orms).
analysis. 6) SPSS
 Type: Sta�s�cal so�ware IV. Prescriptive Analytics
 Usage: Data management, sta�s�cal 1. Op�miza�on Tools: Solver in Excel.
D. Government
analysis, and repor�ng 2. Decision Support Tools: Simple
• Policy Development: Analy�cs supports
 Pros: User-friendly, extensive sta�s�cal simula�on models using
evidence-based policymaking by
capabili�es, strong data manipula�on spreadsheets.
evalua�ng the impact of previous
features 3. Cloud-Based Tools: Salesforce
ini�a�ves.
 Cons: Can be expensive, less flexible Einstein (integrates with business
• Resource Alloca�on: Ensures efficient
compared to some programming opera�ons)
use of funds and resources for public
languages like Python and R
welfare.
7) KNIME
• Crime Preven�on: Predic�ve analy�cs
 Type: Open-source analy�cs pla�orm
helps in crime trend analysis and
 Usage: Data integra�on, repor�ng,
preven�on strategies.
machine learning
Lesson 5:
 Pros: No coding required, modular
 Cons: Performance can be slow with
Overview of Analytics large datasets
Tools and Technologies 8) Weka
 Type: Machine learning and data
Analytics Tools and mining so�ware
Technologies  Usage: Data preprocessing,
1) Microso� Excel classifica�on, regression, clustering,
 Type: Spreadsheet so�ware associa�on rule mining, and
 Usage: Data wrangling, repor�ng, and visualiza�on
visualiza�on  Pros: Extensive library of machine
 Pros: Widely-used, versa�le, many learning algorithms, user-friendly GUI,
built-in func�ons supports various data formats, includes
 Cons: Limited in handling big data tools for data cleaning and prepara�on
2) Python  Cons: Can be less efficient with very
 Type: Programming language large datasets, limited customiza�on
 Usage: Data analysis, machine learning, compared to some programming
web scraping languages
 Pros: Open-source, extensive libraries
(Pandas, NumPy, Matplotlib) Analytics Tools According to
 Cons: Requires programming the Type of Data Analytics
knowledge I. Descriptive Analytics
3) R 1. Spreadsheets: Microso� Excel,
 Type: Programming language Google Sheets.
 Usage: Sta�s�cal analysis, data 2. Visualiza�on Tools: Tableau,
visualiza�on Microso� Power BI.
 4 UNIT 2 DATA FUNDAMENTALS

Lesson 1: • Impact on Decision-Making: which can be used to forecast future

Introduction to Data • Reliable data ins�lls confidence in the events.
decisions being made. V. Driving Iovation
Data o Data is a key driver of innova�on. By
• refers to raw facts and figures collected H. Acceibility analyzing data, organiza�ons can
from observa�ons, measurements, or • Impact on Analysis: Easy access to data iden�fy new opportuni�es, improve
processes. facilitates quicker and more thorough products and services, and create new
• It serves as the founda�on for genera�ng analysis. business models.
insights and decision-making in analy�cs. • Impact on Decision-Making: Accessible VI. Scientific Research
• can take various forms, including numbers, data enables �mely and informed o In science and medicine, data is the
text, images, or even audio and video, decisions. cornerstone of research. It provides the
depending on the context of collec�on. evidence needed to test hypotheses,
I. Granularity validate theories, and advance
Characteristics of Data • Impact on Analysis: Detailed data knowledge.
A. Completene allows for deep, nuanced analysis, while VII. Transparency and Accountability
• Impact on Analysis: Incomplete data aggregated data is useful for high-level o In governance and public
may lead to biased results or missed overviews. administra�on, data promotes
insights. • Impact on Decision-Making: Granular transparency and accountability. It
• Impact on Decision-Making: Decisions data supports specific, targeted allows ci�zens to see how decisions are
made without the full picture can lead decisions, whereas aggregated data made and hold authori�es accountable.
to subop�mal outcomes. aids in strategic planning.
Lesson 2:
B. Consistency J. Security Data TYpes
• Impact on Analysis: Inconsistent data • Impact on Analysis: Secure data
can result in confusion and errors in protects against unauthorized form the founda�on of data analy�cs,
analysis. altera�ons, ensuring data integrity. serving as a cri�cal framework for
• Impact on Decision-Making: • Impact on Decision-Making: Security understanding how data can be
Consistency ensures that decisions are measures ins�ll trust in the data being categorized, analyzed, and u�lized
based on reliable and stable data. used for decision-making. effec�vely. By classifying data into
different types we gain clarity on its
C. Timeline Importance of Data characteris�cs and the appropriate
• Impact on Analysis: Outdated data Data is the backbone of analy�cs, serving as methods for processing and analyzing it.
might not reflect current condi�ons or the raw material for genera�ng insights, By exploring the dis�nct characteris�cs
trends. making predic�ons, and suppor�ng decision- and classifica�ons of data, we can unlock
• Impact on Decision-Making: Timely making. the true power of informa�on and
data ensures that decisions are relevant I. Foundation for Insights transform raw numbers into meaningful
and current. o Data provides the factual basis for narra�ves that drive informed decision
understanding trends, paterns, and making
D. Relevance rela�onships. Without data, analy�cs
E. Impact on Analysis: According to Nature of Data
would lack credibility and depth.
• Irrelevant data can cluter analysis and 1. Quantitative data
II. Informed Decision-Making
obscure important insights. • “Numerical”
o Organiza�ons rely on data-driven
• Impact on Decision-Making: insights to make strategic decisions, • can be counted, measured, and
• Using relevant data helps make focused allocate resources, and iden�fy expressed using numbers.
and effec�ve decisions. opportuni�es. It reduces guesswork  Ex. Height, weight, time
and increases precision. 2. Qualitative data
F. Validity III. Performance Measurement • “Categorical”
• Impact on Analysis: Invalid data can o Data is essen�al for measuring and • descrip�ve and conceptual.
distort the results of the analysis. evalua�ng performance. Organiza�ons • It can be categorized based on traits
• Impact on Decision-Making: Decisions can track their progress against goals, and characteris�cs.
based on invalid data can lead to iden�fy areas for improvement, and  Ex. Gender, color, ice cream flavor,
missteps and inefficiencies. make necessary adjustments. occupation
IV. Understanding Paerns and
G. Reliability Trends
• Impact on Analysis: Reliable data o Analyzing data helps iden�fy paterns
produces consistent and repeatable and trends over �me. This can lead to
analysis results. insights about consumer behavior,
market trends, and social dynamics,
5 UNIT 2 DATA FUNDAMENTALS

According to Continuity of - Quan�ta�ve Data Sources

values - Es�mated 20% of enterprise data I. Based on Coection Method
1. Continuous (Gartner) 1. Primary Data Sources
• “measured” - Numbers, dates, and strings • (First-hand data collec�on)
• Con�nuous data can take infinitely 2. Unstructured Data • These sources involve direct
many values within a range and o�en • Doesn’t have any predefined structure interac�on with respondents or
involves measurement. to it and come in all its diversity of direct recording of data.
• Con�nuous data does not have to be forms  Survey & Ques�onnaire,
whole numbers; it can include - Cannot be displayed in rows, columns, Observa�on, Interview & Focus
frac�onal values. and rela�onal databases Group, Sensors & IoT Devices
 Ex. Height, weight, temperature - Qualita�ve 2. Secondary Data Sources
2. Discrete - Es�mated 80% of enterprise data • (Pre-exis�ng data)
• “counted” (Gartner) • These sources use data that has
• Data that is counted and takes specific, - Images, audio, video, word processing already been collected and
dis�nct values. files, e-mails, spreadsheets published.
• Discrete data is that which is recorded  Research Papers & Reports,
in whole According to Role in Analysis Public Datasets, Web Scraping &
 Ex. No. of books, score, no. of (Functional Relationship) Social Media, APIs & Third-Party
households 1. Independent Variable Databases
• Predictor variable
According to Scale of • The factor that you manipulate or II. Based on Origins
Measurement change in an experiment. 1. Internal Data
1. Nominal • It’s the cause or the condi�on you • (Collected within an organiza�on)
• also called categorical variables control to see how it affects something • Internal data refers to data that
• don’t have a numeric value else. originates from within a company or
• cannot be added, subtracted, divided or • Think of it as the input or the "what organiza�on. It is o�en proprietary
mul�plied. you change" part of the experiment. and used for internal decision-
 Ex. Gender, color ice cream flavor, • It’s what you manipulate or categorize making.
occupation to see how it affects the outcome.  Transactional Data - Sales
2. Ordinal  Ex. Hours studied (used to predict records, invoices, customer
• a property defined by an opera�on test scores). transac�ons
whereby members of a par�cular group 2. Dependent Variable  Operational Data - Inventory
of ranked. • Criterion variable levels, produc�on logs,
• contains things that you can place in • The factor that you measure or observe employee records
order. in response to the independent  Customer Data - Feedback,
 Ex. Educational level, economic variable. complaints, loyalty programs
status, customer satisfaction rating  IT System Logs - Website
• It’s the effect or outcome that you
scale analy�cs, server logs, security
check to see if it changes.
logs
3. Interval • Think of it as the output or the "what
• Interval data is a type of data which is you observe" part of the experiment.
2. External Data
measured along a scale, in which each • It’s what you measure to see if there’s
• (Collected from outside the
point is placed at an equal distance an effect.
organiza�on)
(interval) from one another.  Ex. Test scores (which change based
• External data is gathered from third-
 Ex. Temperature, score on hours studied).
party sources and can provide
4. Ratio
Lesson 3:
addi�onal insights beyond internal
• Has all proper�es of an interval variable data.
and has a clear defini�on of 0 Zero that Data Collection  Public Data - Government
means the absence or none of the thing databases, census data, weather
being measured The process of gathering and analyzing
reports
 Ex. Weight, height informa�on or data from various sources
 Market Data - Compe�tor
in order to address problems/ issues,
pricing, industry reports,
According to Structure of provide answers to queries, assess
economic indicators
Data results, and predict trends and
 Social Media & Web Data -
1. Structured Data probability
Customer sen�ment, online
• Informa�on that is highly organized,
trends, reviews
factual, and to-the-point
 Third-Party Databases & APIs -
- Can be displayed in rows, columns,
Financial data, loca�on data,
and rela�onal databases
demographic insights
6 UNIT 2 DATA FUNDAMENTALS

Primary Data (Collected firsthand) • Iden�fy and delete duplicate records Method 3: Using Condi�onal Forma�ng to
Internal Data External Data that may have been introduced due to Highlight Duplicates
(Within (Outside mul�ple data sources or entry errors. - Select the Column or Range
Organiza�on) Organiza�on) 2. Handle Miing Data - Go to Home> Condi�onal Forma�ng >
Employee surveys, Market surveys, • Deletion Methods: Remove rows or Highlight Cells Rules > Duplicate Values
customer compe�tor columns with missing values (only if - Choose a forma�ng style (e.g., red fill) and
feedback, system interviews, field minimal impact). click OK.
logs, IoT data observa�ons, web • Imputation: Fill missing values using
scraping mean, median, mode, or advanced
Secondary Data (Pre-exis�ng data) techniques like regression or machine
Internal Data External Data learning.
(Within (Outside 3. Standardize Data Formats
Organiza�on) Organiza�on) • Standardize formats for dates, text,
Historical sales, Government numbers, and categorical labels (e.g.,
archived reports, reports, research “M” vs. “Male”).
past security logs papers, third-party 4. Correct Data Entry Errors
APIs, social media • Iden�fy and fix typos, incorrect values,
analy�cs and misplaced decimal points.
5. Detect and Handle Outliers
 Primary Data is useful when you need • Use sta�s�cal techniques and
firsthand, specific, and accurate data. visualiza�on to iden�fy and decide
 Secondary Data is valuable for whether to keep, modify, or remove
benchmarking, trend analysis, and cost- outliers.
effec�ve insights. 6. Normalize and Scale Data II. Handling Miing Data
 Internal Data helps in improving • Convert numerical data to a consistent Identify Missing Data
business opera�ons and customer scale using normaliza�on or
sa�sfac�on. standardiza�on for beter analysis. Method 1: Use Filters
 External Data helps in understanding 7. Validate and Verify Data Integrity - Select your dataset.
market trends and compe�tor • Cross-check with source data and run - Click Data→ Filter.
strategies. valida�on checks to ensure data - Click the dropdown arrow on each column
accuracy and completeness. and filter for blanks.
Lesson 4:

Data Preparation and Lesson 5:

Cleaning Data Cleaning Using

is the process of transforming raw data
excel
into a structured and usable format for
I. Removing Duplicate Data
analysis It involves handling missing
Method 1: Using the "Remove Duplicates"
values, correc�ng inconsistencies,
Tool
standardizing formats, and ensuring data
quality

Steps in Data Preparation

1. Acquiring Data Method 2: Using the UNIQUEFunc�on (Excel
• Collec�ng data from various sources 365 & Excel 2019)
2. Exploring Data - The UNIQUE func�on returns a list of
• Reviewing the dataset structure, unique values in a list or range.
iden�fying missing values, duplicates, - UNIQUE(array, [by_col], [exactly_once])
and inconsistencies. Method 2: Condi�onal Forma�ng
3. Cleaning Data - Select your data range.
• Fixing or removing errors - Go to Home→ Condi�onalForma�ng→
4. Transform Data New Rule.
• Conver�ng data into a usable format - Choose Format only cells that contain →
Set Blanks.
Key Steps in Data Cleaning - Click Format .Apply a color to highlight
1. Remove Duplicate Data missing values.
7 UNIT 2 DATA FUNDAMENTALS

III. Standardizing Data Formats

Standardizing Number Formats
(date, currency, percentage etc.)
- Select column → Press Ctrl + 1 → Choose a
Category under the Number tab.
or
- Go to the Home tab. In the Number group,
locate the Number Format dropdown.
or
- Select Column , Right click → Format

Standardizing Text Case(ensuring

consistent capitalization)
- Use func�ons =PROPER() (Proper Case),
=UPPER() (All Caps), =LOWER() (All
Lowercase)

Removing Extra Spaces & Symbols

- Use func�ons =TRIM() to remove spaces
and Find & Replace (Ctrl + H) to remove
unwanted characters

Method 3: Use COUNTBLANKFunc�on

- =COUNTBLANK(range)
- This counts the number of blank cells in a
range.

Common Techniques for Handling

Missing Data
1. Fill with the Previous or Next Value (for
�me-series data).
2. Fill with an Average, Median, or Most
Frequent Value (for
numerical/categorical data).
3. Fetching missing values from a
reference table using VLOOKUP
4. Remove Rows with Missing Data (only if
necessary).