Data Science Fundamentals: Quiz 1

Study Notes

Data science is a collection of techniques extracting value from data.
Techniques have roots in statistics, machine learning, visualization, logic, and computer science.
Data science uses patterns, connections, and relationships in data.
Data science is also known as knowledge discovery, machine learning, predictive analytics, and data mining.
Underlying methods are decades or centuries old.
Data science methods are evidence-based and built on empirical and historical observations.

Increases efficiencies, manages costs, identifies market opportunities, and boosts market advantage.
Enables mining of large data sets (structured and unstructured) to identify patterns and insights.
An interdisciplinary field combining statistics, computer science, predictive analytics, machine learning, and new technologies to gain insights from big data.

Artificial intelligence, machine learning, and data science are related.
Often used interchangeably in media and business.
Artificial intelligence gives machines the capability to mimic human behavior (e.g., facial recognition, automated driving, mail sorting).
Machines can exceed human capabilities in some areas but also have limitations ("artificial stupidity") when not properly programmed or with incomplete data (e.g., self-driving car misunderstanding detours).
Machine learning is a sub-field or tool of AI, giving machines the ability to learn from experience.

Experience for machines is data.
Training data is used to teach machines.
Machine learning algorithms take known input and output (training data) to create a program converting input to output.
Programs (instructions) transform input signals into output signals following predetermined rules and relationships.

Removing abusive content on social media (using examples of abusive and non-abusive posts).

This cycle encompasses these steps: Capture, Prepare and Maintain, Preprocess or Process, Analyze, and Communicate.
Capture: Gathering raw data from various sources.
Prepare and Maintain: Transforming raw data to a consistent format, cleansing, deduplicating, and reformatting. ETL tools are often used.
Preprocess or Process: Analyzing biases, patterns, ranges, and distributions in the data, preparing it for use in machine learning and predictive analytics.
Analyze: Performing statistical analysis, predictive analytics, regression, machine learning, or deep learning algorithms to extract insights.
Communicate: Presenting insights through reports, charts, and visualizations to stakeholders.

Business Acumen: Understanding of business domain, business strategy, problem solving, communication, presentation, inquisitiveness.
Technology Expertise: Good knowledge of databases, programming languages, open-source tools, data warehousing, data mining, and visualization tools.
Mathematical Expertise: Knowledge of mathematics, statistics, artificial intelligence, machine learning, pattern recognition, and natural language processing.

Analyze business data for meaningful insights.
Solve business problems through steps including data collection and analysis, determining the problem, defining the correct variables and data sets, gathering structured and unstructured data from disparate sources, processing raw data into analyzable data, analysis and identifying trends and patterns, and interpreting data to identify opportunities and solutions.
Preparing insights and results for stakeholders and communicating those insights.