Introduction to Data Science

Introduction to Data Science

• Data science relies on data as the foundation for analysis.
• Data, in the context of data science, encompasses raw facts, figures, observations, and information.
• Data is collected, stored and analyzed to achieve insights, support informed decision-making and enhance various applications.
• Diverse sources include sensors, surveys, social media, business transactions and scientific experiments.
• Data examples: the average customer age in a dataset is 35, product sales increased 20% last quarter or 80% of users prefer a mobile app for online shopping.

What is Data: Raw Facts

• Raw facts are specific pieces of information deduced from data analysis.
• Examples of raw facts: the average age of customers within a dataset is 35 years old, product sales have increased by 20% compared to the previous quarter, or 80% of users prefer mobile apps for online shopping

What is Data: Figures

• Figures usually represent visualizations of data, illustrating patterns, trends and relationships within a dataset.
• Example: a scatter plot used to illustrate the relation between study hours and exam scores.
• Scatter plots visually show the connection between these two factors.

What is Data: Observations

• Observations refer to specific instances or data points gathered from experiments, surveys, studies or other research activities.
• Observations form the bedrock for data analysis.
• Example: Observations related to customer satisfaction surveys contain customer ID, date of purchase, satisfaction rating (on a scale of 1-5), and comments.

Data Sources

• Data can come from various sources, including computers, mobile devices, cameras, sensors, wearables, social media interactions, saved files, photos and queries.
• Even simple actions like getting directions generate data.

Types of Data

• Data exists in numerous forms: numbers, text, images, audio, video and more.
• In data science, data types are categorized as structured and unstructured.

Structured Data

• Structured data is organized in a defined format, which algorithms easily process.
• This format facilitates searching and analysis for both humans and machines.
• Structured data is suitable for quantitative analysis.
• It's often tabular, employing rows and columns that explicitly define data attributes.

Structured Data Examples

• Databases: Information stored in relational databases, organized as tables with rows and columns.

• Examples are: MySQL, PostgreSQL.

• Spreadsheets: Data arranged in rows and columns in spreadsheets such as Microsoft Excel or Google Sheets.

• CSV (Comma-Separated Values) Files: Plain text files where values are separated by commas and each line represents a new record.

• JSON (JavaScript Object Notation) Data: Data presented using key-value pairs organized in a hierarchical structure.

• XML (eXtensible Markup Language) Data: Data presented using tags in a hierarchical structure suitable for complex and nested relationships.

Unstructured Data

• Unstructured data lacks a predefined data model or format.
• It's less easily organized into tables than structured data.
• It comes in diverse formats like text documents, images, audio files, videos, emails and social media posts.

Unstructured Data Examples

• Text Data: Textual information from various sources such as articles, books, emails, social media posts and customer reviews.
• Image Data: Photographs, diagrams, satellite images and other visual data.
• Audio Data: Recordings of voice conversations, music, podcasts and other audio sources.
• Video Data: Moving images captured by cameras or videos.
• Social Media Feeds: Data from social media platforms including posts, comments and multimedia content.
• Web Pages: Content from websites, articles, blogs and forum posts.

Structured vs. Unstructured Data Comparison

• Structured data is readily displayed in rows and columns of relational databases.
• Structured data comprises numbers, dates and strings. It represents approximately 20% of enterprise data.
• Structured data utilizes less storage and offers easier management and protection using legacy solutions.
• Unstructured data can't be displayed in rows and columns as it lacks a defined format.
• Unstructured data includes images, audio, video, emails, and spreadsheets. It represents approximately 80% of enterprise data.
• Unstructured data requires more storage and is more complex to manage and protect using legacy solutions.

Big Data

• Big data is more extensive than traditional data.
• It's recognized by its diversity (text, images, audio etc.), velocity (real-time retrieval/computation), and volume (measured in tera, peta and exabytes). This massive data is dispersed across a computer network.

Quantitative vs Qualitative Data

• Quantitative data includes numerical measurements. It's expressed using numbers.
• Qualitative data relies on observations and descriptions. It's not easily expressed in numbers.

Quantitative Data

• Nature: Numerical data consisting of measurable values.
• Measurement: Using precise and standardized procedures. Units of measurement are included (e.g., kilograms).
• Analysis: Statistical methods like mean, median, standard deviation and visualization methods such as histograms, bar graphs, line graphs and scatter plots.
• Example: Age (25 years), weight (70kg), revenue ($5,000) or quantity (100 units).

Qualitative Data

• Nature: Descriptive, focusing on qualities and characteristics. Non-numerical, relying on categorical or textual information.
• Measurement: Subjective, based on insights, opinions, attitudes and often presented in a narrative format. Categorized into groups or classes.
• Analysis: Narrative analysis, involving the examination of stories, texts and other qualitative data for valuable insights.
• Example: Marital status (Married), customer feedback ("Satisfied"), or colors (red, blue, green).

This quiz covers the foundational concepts of data science, including the role of data, definitions, and specific types of data such as raw facts and figures. Test your understanding of how data is collected, analyzed, and utilized to derive insights and inform decisions in various applications.