CISSE Advanced Level
Big Data Analytics cheat sheet
Every chapter's key points, formulas and definitions on one page, for the last week before the exam. Each heading links to the chapter that explains it.
Introduction to big data analytics
- Big data is defined by five V's: Volume, Velocity, Variety, Veracity and Value, each a distinct reason ordinary data-processing tools become inadequate.
- Big data analytics applies statistical, computational and machine learning techniques to data at a volume, velocity and variety beyond what conventional analysis on a single machine can handle.
- AI, machine learning and deep learning are nested: every deep learning system is machine learning and every machine learning system is AI, but a rule-based expert system is AI with no learning from data.
- Each core challenge has a matching answer: tools built for scale, systematic cleaning for veracity, rigorous validation against spurious patterns, and a legal and ethical framework for privacy.
- Python dominates the field because of its large, mature ecosystem of data science libraries, a simple syntax and strong interoperability with large-scale distributed systems.
Key terms
- Volume
- The sheer scale of data generated and stored, often too large for a single machine's memory or a traditional relational database to process efficiently.
- Velocity
- The speed at which new data is generated and must be processed, often in near real time (a stream of sensor readings, social media posts).
- Variety
- Data arrives in many forms at once, structured (tables), semi-structured (JSON, logs) and unstructured (text, images, video).
- Veracity
- The trustworthiness and quality of the data, since big data is frequently noisy, incomplete or inconsistent (developed in full in the chapter on exploring datasets).
- Value
- The whole point of the exercise: data has no worth until analysis turns it into a decision or insight, the value this entire paper's techniques exist to extract.
- Artificial intelligence (AI)
- The broad field of building systems that perform tasks normally requiring human intelligence, reasoning, perception, decision-making.
- Machine learning (ML)
- A subfield of AI: systems that improve their performance on a task by learning patterns from data, rather than being explicitly programmed with fixed rules, the subject of the machine learning chapters.
- Deep learning
- A further subfield of machine learning using artificial neural networks with many layers, developed in full in the chapter on deep learning.
The rest of the sheet covers
- Mathematics for big data analytics
- Exploration of massive datasets in Python
- The machine learning pipeline
- Unsupervised machine learning (clustering) in Python
- Supervised machine learning (classification) in Python
- Association rule mining in Python
- Deep learning implementation in Python
- Natural language processing in Python
- Big data visualisation in Python
- The law and ethics in big data analytics
- Worked examples
- Exam technique and revision
- Glossary