Google, Microsoft, IBM and AWS are just some of the tech behemoths taking on machine learning, creating APIs and developing a number of sophisticated deep learning frameworks.

Looking for a new job? Browse Techworld's jobs channel.

© iStock/iLexx
© iStock/iLexx

As new areas of technology are exploited and pulled into the mainstream, the demand for skilled workers begins to rise.

This is our guide to getting started with machine learning.

What is machine learning?

First coined in 1959 by Arthur Samuel - a computer scientist at IBM at the time - "Machine Learning" essentially enables computers to learn without being directly programmed.

Machine learning (ML) is fundamentally the application of AI that we recognise today, for example, machines performing 'smart' tasks.

Often confused with artificial intelligence (AI), ML is a little different. AI is the general term for connected and 'smart' machines, and the tasks they perform.

So while ML is the science behind machines performing tasks without command, AI refers to building machines capable of intelligent behaviour.

What skills are required?

The path to becoming a machine learning engineer isn't an easy one.

You'll have to already have a good base of technical knowledge and an analytical/mathematical brain to match.

To start, you should brush up on your maths skills. Machine learning is essentially applied statistics and mathematics.

Having a good grasp of applied maths and how statistics work is very important, as it will help you pick up algorithmic sequences easier. 

Get to the library and read about statistics and probability models, maybe even look into quadratic programming and partial differential equations as all of this will help you later when you need to create self-learning algorithms.

Good reads include An Introduction to Statistical Learning by Robert Tibshirani and Trevor Hastie or Machine Learning with R by Brett Lantz.

You should also have knowledge of a database, statistical programming language. Both Python and R programming languages are great choices, as they are able to process large data sets and statistics. 

Python is extremely readable and easier to learn, however, R provides a more complete statistical language. Most data science roles involve either Python or R, as well as C++ for its scripting capabilities and ability to speed up code.

If you'd like to learn R, check out: What is R and why should I learn it?.

In machine learning, you need to modify and rework the structure of datasets and you'll most likely use Hadoop HBase to store it, so learning Java would be advantageous (seeing as Hadoop is built on Java).

Ideally, you'll have a degree or in-depth knowledge of computer science, with solid knowledge of computing structures, data structures like stacks, queues and multi-dimensional arrays, as well as data modelling (estimating the underpinning structure of a given dataset).

Finally, you'll need some experience with Unix tools. These include cat, grep, find, awk, sed, sort and cut. Given the majority of machine learning processing runs primarily only on the Linux-based machines, Unix knowledge is key.

Which industries are using machine learning?

While machine learning is firmly in the pipeline for most businesses in some way, adoption is still noteworthy, and definitely not the norm just yet.

The financial sector has seen particular uptake of ML, with banks finding numerous uses for the algorithmic technology.

For example, banks can use ML to combat fraud by combing through huge transactional data sets to spot unusual behaviour.

In fact, countless things in the financial sector could become reliant on ML, from credit card applications and algorithmic training to AI-managed funds and chatbots.

But it doesn't stop there. Healthcare is a something that could benefit hugely from ML.

Machines will be able to spot and track life-threatening illnesses, just like spotting an anomaly on a document in the financial services.

This could save health services millions, if not billions each year. It should also speed up the rate of detection, thus inevitably saving lives.

Transport will also see the impact of machine learning, and to an extent, it already has. Examples of which can be seen in railway track sensors and sensors to monitor road surface degradation.

With machine learning, you're able to expose vehicles to millions of potential scenarios and make sure the computer in the car, bus or truck learns to acts in a certain way. 

For a full list of machine learning uses, read this.

Machine learning online courses

If you want to brush up on your skills, or just want to dip your toe in the water, you should try an online course. 

There's a myriad of online training providers that offer courses from as little as £15 for beginners. 

From Udemy:

Data Science and Machine Learning Bootcamp with R
Machine Learning and Data Science Essentials with Python & R
Machine Learning A-Z: Hands-On Python & R In Data Science

From Udacity:

Become a machine learning engineer (nanodegree)

From FutureLearn: 

Big Data: Statistical Inference and Machine Learning
Advanced Machine Learning

From Coursera:

Machine Learning Foundations: A Case Study Approach
Applied Data Science with Python 
Practical Machine Learning