BI unit -1

Domain Label: collage courses Published: July 17, 2026 Creator: Sumit Haldar Est. Read: 4 min read

Business Intelligence

What is Data Analytics?

Data Analytics is the process of collecting, organizing, cleaning, analyzing, and interpreting raw data into meaningful information that supports better decision-making.

Hierarchy of Data Processing:

Data → Information → Knowledge → Wisdom

Note: Unwanted data is referred to as Noise in computer science.

1. Data Science Life Cycle

Business Understanding: Defining objectives and business requirements.
Data Collection: Gathering raw data from various sources.
Data Processing and Cleaning: Handling errors and organizing data.
Data Analysis: Performing exploratory analysis and extracting patterns.
Data Modeling: Building predictive and machine learning models.
Model Deployment: Implementing the model into production environments.

Note: GIGO principle applies — Garbage In, Garbage Out.

2. Data Collection Methods & Types

A. Data Collection Methods

Primary Data: First-hand collected data (e.g., surveys, interviews).
Secondary Data: Previously collected data (e.g., existing databases, sentiment analysis datasets).

B. Types of Data

Data Type	Description	Examples / Formats
Structured Data	Highly organized data stored in tabular format.	Excel (.xlsx), SQL Databases
Semi-Structured Data	Does not conform to rigid tables, but uses organizational tags.	JSON, XML
Unstructured Data	Data with no predefined conceptual structure.	Photos, Audio, Video, Instagram Posts
Time Series Data	Data points indexed in time order.	Stock prices, weather updates

Qualitative Data: Non-numerical data capturing opinions, experiences, concepts, and feedback.

Quantitative Data: Numerical data used to measure variables and uncover statistical patterns.

3. Data Quality Problems & Cleaning Process

Common Data Quality Issues

Missing values
Duplicate data
Incorrect data
Inconsistent data
Outliers
Formatting errors

Data Cleaning Steps

Identify Errors
Handle Errors
Validate the Data
Save Clean Data
Data Transformation
Data Integration

4. Web Scraping Cycle & Data Integration

Web Scraping Life Cycle:

        Input URLs → Request Data → Load HTML → Parse HTML → Extract Data → Clean Data → Paginate → Save Data
    

Types of Data Integration:

Manual Data Integration
Middleware Data Integration
Data Warehousing (Tools: Apache Kafka, Talend, MuleSoft)

5. Exploratory Data Analysis (EDA)

EDA is the process of understanding the structure and characteristics of data before formal modeling or building Machine Learning algorithms.

Main Objectives of EDA:

Understand the structure of data.
Summarize main characteristics.
Detect missing values and outliers.
Understand feature distributions.
Discover relationships between variables.
Generate hypotheses for further analysis.

Types of EDA:

Univariate EDA: Analyzes one variable at a time.
Bivariate EDA: Analyzes the relationship between two variables.
Multivariate EDA: Analyzes relationships among three or more variables.

Descriptive Statistics in EDA:

Mean / Median / Mode Min / Max Standard Deviation Variance

EDA & Analytics Workflow

        Raw Data → Data Cleaning → Data Understanding → Perform Statistical Analysis → Visualize Data & Find Patterns → Prepare ML Model
    

6. Data-Driven Decision Making & Types of Analytics

Data-Driven Decision Making (DDDM) involves using facts, metrics, and data to guide strategic business decisions aligned with goals and initiatives.

Descriptive Analytics

Summarizes historical data and past performance through data aggregation and mining.

Diagnostic Analytics

Examines data deeper to answer why things happened, uncovering root causes and correlations.

Predictive Analytics

Forecasts future trends and possibilities using historical data patterns and machine learning.

Real-Time Analytics

Provides immediate insights on incoming data using streaming analytics, event processing, and live dashboards.