3.7

Artificial Intelligence

Artificial intelligence is built around several key concepts, including:

knowledge representation

natural language processing

neural networks

machine learning

planning.

Big Data

The field of big data intersects with AI because machines need a lot of data to train their algorithms.

Data is classically quantified by the four Vs:

volume – the amount of data

velocity – the speed at which the data is generated

variety – the different types of data

veracity – the trustworthiness of the data in terms of accuracy.

he EU’s General Data Protection Regulation (GDPR) was an important step in defending the individual’s rights over their data and stopping abuse by corporations.

Data analytics

Data analytics is used to observe trends and make predictions.

MATLAB is a key tool for data science and statistical analysis.

The Apache SparkTM suite of libraries for processing and performing ETL (extract, transform, load) operations on Hadoop data includes:

Structured Query Language (SQL) – the industry standard for searching relational databases

Spark Streaming – high-throughput processing of live data streams

MLib – machine learning and statistical analysis

GraphX – graphing

AI tools for businesses IBM Watson, Microsoft Azure, Amazon Web Services, Google Data Studio, Oracle, and SAP Analytics Cloud.

Fundamentals of AI

AI and knowledge

AI is a branch of computer science which aims to create intelligent machines

This involves developing computer programs to complete tasks that would otherwise require human intelligence

Knowledge is a core part of AI

AI aims to provide machines with knowledge capacity and algorithms to perform functions such as learning, perception, problem-solving, understanding language and logical reasoning without human intervention.

AI

Machine Learning

Deep Learning

natural language processing (NLP)

knowledge representation and reasoning

optimisation methods and decision making

Knowledge representation is concerned with representing information in a form that a computer algorithm can easily navigate and interpret

Alan Turing, the pioneer of AI

Turing Test

The final question for AI is: can a computer be conscious of its own existence? This, and other more philosophical questions about AI will be covered in Unit 6.

Case Studies

AI in advertising

AI in medicine

use in diagnostic medicine

It can identify patterns and detect irregularities better than a human

Using AI to diagnose disease leaves clinicians with more time to spend treating patients.

Precision medicine is the customisation of healthcare, with medical diagnosis and treatment that is tailored to the individual patient.

Big data and healthcare

To make sense of big data, data analytics is required, including:

predictive analytics – forecasting behaviours and future developments

machine learning – using algorithms to analyse large data sets.

data mining – data sets are searched for patterns and relationships

Advanced patients care

Improved operational efficiency

Improved Treatment

Electronic health records will help to obtain demographic and medical data such as lab tests, clinical data, diagnosis and medical conditions.

Big data can be used as part of a business intelligence strategy to analyse historical patient admission rates and staff efficiency.

Big data can help to identify unknown correlations, hidden patterns and insights by examining significant amounts of data.