Deep learning is machine learning, and machine learning is artificial intelligence. But how do they fit together (and how do you get started learning)?
Even if you’re not involved in the world of data science, you’ve probably heard the terms artificial intelligence (AI), machine learning, and deep learning thrown around in recent years. Sometimes, they’re even used interchangeably. While related, each of these terms has its own distinct meaning, and they're more than just buzzwords used to describe self-driving cars.
In broad terms, deep learning is a subset of machine learning, and machine learning is a subset of artificial intelligence. You can think of them as a series of overlapping concentric circles, with AI occupying the largest, followed by machine learning, then deep learning. In other words, deep learning is AI, but AI is not deep learning.
Thanks to pop culture depictions from 2001: A Space Odyssey to The Terminator, many of us have some conception of AI. Oxford Languages defines AI as “the theory and development of computer systems able to perform tasks that normally require human intelligence.” Britannica offers a similar definition: “the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings.”
Machine learning and deep learning are both types of AI. In short, machine learning is AI that can automatically adapt with minimal human interference. Deep learning is a subset of machine learning that uses artificial neural networks to mimic the learning process of the human brain.
Take a look at these key differences before we dive in further.
|Machine learning||Deep learning|
|A subset of AI||A subset of machine learning|
|Can train on smaller data sets||Requires large amounts of data|
|Requires more human intervention to correct and learn||Learns on its own from environment and past mistakes|
|Shorter training and lower accuracy||Longer training and higher accuracy|
|Makes simple, linear correlations||Makes non-linear, complex correlations|
|Can train on a CPU (central processing unit)||Needs a specialized GPU (graphics processing unit) to train|
Natural language processing (NLP) is another branch of machine learning that deals with how machines can understand human language. You can find this type of machine learning with technologies like virtual assistants (Siri, Alexa, and Google Assist), business chatbots, and speech recognition software.
At its most basic level, the field of artificial intelligence uses computer science and data to enable problem solving in machines.
While we don’t yet have human-like robots trying to take over the world, we do have examples of AI all around us. These could be as simple as a computer program that can play chess, or as complex as an algorithm that can predict the RNA structure of a virus to help develop vaccines.
Before the development of machine learning, artificially intelligent machines or programs had to be programmed to respond to a limited set of inputs. Deep Blue, a chess-playing computer that beat a world chess champion in 1997, could “decide” its next move based on an extensive library of possible moves and outcomes. But the system was purely reactive. For Deep Blue to improve at playing chess, programmers had to go in and add more features and possibilities.
For a machine or program to improve on its own without further input from human programmers, we need machine learning.
Machine learning refers to the study of computer systems that learn and adapt automatically from experience, without being explicitly programmed.
With simple AI, a programmer can tell a machine how to respond to various sets of instructions by hand-coding each “decision.” With machine learning models, computer scientists can “train” a machine by feeding it large amounts of data. The machine follows a set of rules—called an algorithm—to analyze and draw inferences from the data. The more data the machine parses, the better it can become at performing a task or making a decision.
Here’s one example you may be familiar with: Music streaming service Spotify learns your music preferences to offer you new suggestions. Each time you indicate that you like a song by listening through to the end or adding it to your library, the service updates its algorithms to feed you more accurate recommendations. Netflix and Amazon use similar machine learning algorithms to offer personalized recommendations.
In 2011, IBM Watson beat two Jeopardy champions in an exhibition match using machine learning.
Watson’s programmers fed it thousands of question and answer pairs, as well as examples of correct responses. When given just an answer, the machine was programmed to come up with the matching question. If it got it wrong, programmers would correct it. This allowed Watson to modify its algorithms, or in a sense “learn” from its mistakes.
By the time Watson faced off against the Jeopardy champions, in a matter of seconds, it could parse 200 million pages of information and generate a list of possible answers, ranked by how likely they were to be right—even if it had never seen the particular Jeopardy clue before.
Where machine learning algorithms generally need human correction when they get something wrong, deep learning algorithms can improve their outcomes through repetition, without human intervention. A machine learning algorithm can learn from relatively small sets of data, but a deep learning algorithm requires big data sets that might include diverse and unstructured data.
Think of deep learning as an evolution of machine learning. Deep learning is a machine learning technique that layers algorithms and computing units—or neurons—into what is called an artificial neural network. These deep neural networks take inspiration from the structure of the human brain. Data passes through this web of interconnected algorithms in a non-linear fashion, much like how our brains process information.
AlphaGo was the first program to beat a human Go player, as well as the first to beat a Go world champion in 2015. Go is a 3,000-year-old board game originating in China and known for its complex strategy. It’s much more complicated than chess, with 10 to the power of 170 possible configurations on the board.
The creators of AlphaGo began by introducing the program to several games of Go to teach it the mechanics. Then it began playing against different versions of itself thousands of times, learning from its mistakes after each game. AlphaGo became so good that the best human players in the world are known to study its inventive moves.
The latest version of the AlphaGo algorithm, known as MuZero, can master games like Go, chess, and Atari without even needing to be told the rules.
The term “big data” refers to data sets that are too big for traditional relational databases and data processing software to manage. Businesses are generating unprecedented amounts of data each day. Deep learning is one way to derive value from that data. Read: What is Big Data? A Layperson's Guide
If this introduction to AI, deep learning, and machine learning has piqued your interest, AI for Everyone is a course designed to teach AI basics to students from a non-technical background.
For more advanced knowledge, start with Andrew Ng’s Machine Learning course for a broad introduction to the concepts of machine learning. Next, learn to build intelligent applications with the Machine Learning Specialization. Finally, build and train artificial neural networks in the Deep Learning Specialization.
When you’re ready, start building the skills needed for an entry-level role as a data scientist with the IBM Data Science Professional Certificate, and further practice your skills with these portfolio-ready, hands-on projects.
Machine Learning Pipelines with Azure ML Studio: Build an end-to-end machine learning pipeline using adult income census data.
Detecting COVID-19 with Chest X-Ray using PyTorch: Train a residual neural network (RNN) using a radiography dataset.
Fake News Detection with Machine Learning: Train a deep learning model to detect fake news from a news corpus.
Sentiment Analysis with Deep Learning using BERT: Explore the concepts of natural language processing (NLP) as you analyze a dataset for sentiment analysis.
Machine learning typically falls under the scope of data science. Having a foundational understanding of the tools and concepts of machine learning could help you get ahead in the field (or help you advance into a career as a data scientist, if that’s your chosen career path).
Machine learning is a field that’s growing and changing, so learning is an ongoing process. Depending on your background and how much time you can devote to learning, it might take you a few weeks, a few months, or a year to build a strong foundation in machine learning. Here are some tips for rising to the challenge.
The technical skills and concepts involved in machine learning and deep learning can certainly be challenging at first. But if you break it down using the learning pathways outlined above, and commit to learning a little bit everyday, it’s totally possible. Plus, you don’t need to master deep learning or machine learning to begin using your skills in the real world.
Deep learning and machine learning as a service platforms mean that it’s possible to build models, as well as train, deploy, and manage programs without having to code. While you don’t necessarily need to be a master programmer to get started in machine learning, you might find it helpful to build basic proficiency in Python.
Yes. The average base pay for a machine learning engineer in the US is $124,878, as of December 2021 . According to a December 2020 study by Burning Glass, demand for AI and machine learning skills is projected to grow by 71 percent over the next five years. The same study reports a $14,175 salary premium associated with these skills .
1. Glassdoor. "Machine Learning Engineer Salaries, https://www.glassdoor.com/Salaries/machine-learning-engineer-salary-SRCH_KO0,25.htm." Accessed April 22, 2022.
2. Burning Glass. "Skills of Mass Disruption: Pinpointing the 10 Most Disruptive Skills in Tech, https://www.burning-glass.com/wp-content/uploads/2020/12/Skills-of-Mass-Disruption-Report.pdf." Accessed April 22, 2022.
This content has been made available for informational purposes only. Learners are advised to conduct additional research to ensure that courses and other credentials pursued meet their personal, professional, and financial goals.