LM 11: Introduction to Big Data Techniques

Question 1

What are the three Vs that Big Data is characterised by and when is a Fourth V relevant?

Answer 1

Volume, Velocity and variety.

Fourth V is relevant when Big Data is used for inference or prediction: Veracity

Question 2

What are the main sources of alternative data?

Answer 2

Individuals, business processes and sensors

Question 3

What is the definition of AI?

Answer 3

Computer systems capable of performing tasks that traditionally required human intelligence at levels comparable to those of human beings

Question 4

Machine Learning is defined as?

Answer 4

Extracts knowledge from large amounts of data by learning from known examples and then generating structure or predictions.

Question 5

What are the 3 main types of ML?

Answer 5

Supervised learning, unsupervised learning and deep learning

Question 6

What is Natural Language Processing (NLP)?

Answer 6

Application of text analytics that uses insight into the structure of human language to analyse and interpret text and voice based data

Question 7

What are traditional sources of big data?

Answer 7

Stock exchanges; companies and governments

Question 8

What are non-traditional data types?

Answer 8

Electronic devised, social media, sensor networks and company exhaust

Question 9

What is Veracity of data?

Answer 9

Relates to the credibility and reliability of different data sources

Question 10

What is structured data?

Answer 10

Organised into tables and commonly stored in a database where each field represents the same type of info

Question 11

What is unstructured data?

Answer 11

Disparate, unorganised data that cannot be organised in tubular form

Question 12

What is semi-structured data?

Answer 12

Has elements of both structured and unstructured data

Question 13

What is an example of unstructured data?

Answer 13

Voice recordings

Question 14

What is the Internet of Things IoT?

Answer 14

Data generated by smart devices about a wide range of information. E.g. Smart building

Question 15

What is Overfitting?

Answer 15

When data discovers false relationships that are unsubstantiated

Question 16

What is Underfitting?

Answer 16

When ML treats important parameters as if they are noise and discards them, making the data set too simplistic

Question 17

What is the key basis of supervised learning?

Answer 17

The data is based on labelled training data

Question 18

What type of data is given in unsupervised learning?

Answer 18

Not given labelled data but instead are given only data from which the algorithm seeks to describe the data and its structure

Question 19

What is deep learning?

Answer 19

Neural networks with many hidden layers to perform multistage non linear data processing to identify patterns

Question 20

What are the the 5 Data Processing methods?

Answer 20

Capture
Curation
Storage
Search
Transfer

Question 21

What is Data Visualisation?

Answer 21

Refers to how the data will be formatted, displayed and summarised in graphical form

Question 22

How do Text analytics work?

Answer 22

Use computer programs to analyse and derive meaning typically from large unstructured text or voice based datasets

Question 23

What is a main characteristic of Big data?

Answer 23

It involved formats with diverse structures