Machine learning is a subset of artificial intelligence (AI) in which computers learn from data and improve with experience without being explicitly programed.
Machine learning definition in detail
Machine learning is a subset of artificial intelligence (AI). It is focused on teaching computers to learn from data and to improve with experience – instead of being explicitly programmed to do so. In machine learning, algorithms are trained to find patterns and correlations in large data sets and to make the best decisions and predictions based on that analysis. Machine learning applications improve with use and become more accurate the more data they have access to.
Applications of machine learning are all around us –in our homes, our shopping carts, our entertainment media, and our healthcare.
How is machine learning related to AI?
Machine learning – and its components of deep learning and neural networks – all fit as concentric subsets of AI. AI processes data to make decisions and predictions. Machine learning algorithms allow AI to not only process that data, but to use it to learn and get smarter, without needing any additional programming. Artificial intelligence is the parent of all the machine learning subsets beneath it. Within the first subset is machine learning; within that is deep learning, and then neural networks within that.
Diagram of the relationship between AI and machine learning
What is a neural network?
An artificial neural network (ANN) is modeled on the neurons in a biological brain. Artificial neurons are called nodes and are clustered together in multiple layers, operating in parallel. When an artificial neuron receives a numerical signal, it processes it and signals the other neurons connected to it. As in a human brain, neural reinforcement results in improved pattern recognition, expertise, and overall learning.
What is deep learning?
This kind of machine learning is called “deep” because it includes many layers of the neural network and massive volumes of complex and disparate data. To achieve deep learning, the system engages with multiple layers in the network, extracting increasingly higher-level outputs. For example, a deep learning system that is processing nature images and looking for Gloriosa daisies will – at the first layer – recognize a plant. As it moves through the neural layers, it will then identify a flower, then a daisy, and finally a Gloriosa daisy. Examples of deep learning applications include speech recognition, image classification, and pharmaceutical analysis.
How does machine learning work?
Machine learning is comprised of different types of machine learning models, using various algorithmic techniques. Depending upon the nature of the data and the desired outcome, one of four learning models can be used: supervised, unsupervised, semi-supervised, or reinforcement. Within each of those models, one or more algorithmic techniques may be applied – relative to the data sets in use and the intended results. Machine learning algorithms are basically designed to classify things, find patterns, predict outcomes, and make informed decisions. Algorithms can be used one at a time or combined to achieve the best possible accuracy when complex and more unpredictable data is involved.
How the machine learning process works
What is supervised learning?
Supervised learning is the first of four machine learning models. In supervised learning algorithms, the machine is taught by example. Supervised learning models consist of “input” and “output” data pairs, where the output is labeled with the desired value. For example, let’s say the goal is for the machine to tell the difference between daisies and pansies. One binary input data pair includes both an image of a daisy and an image of a pansy. The desired outcome for that particular pair is to pick the daisy, so it will be pre-identified as the correct outcome.
By way of an algorithm, the system compiles all of this training data over time and begins to determine correlative similarities, differences, and other points of logic – until it can predict the answers for daisy-or-pansy questions all by itself. It is the equivalent of giving a child a set of problems with an answer key, then asking them to show their work and explain their logic. Supervised learning models are used in many of the applications we interact with every day, such as recommendation engines for products and traffic analysis apps like Waze, which predict the fastest route at different times of day.
What is unsupervised learning?
Unsupervised learning is the second of the four machine learning models. In unsupervised learning models, there is no answer key. The machine studies the input data – much of which is unlabeled and unstructured – and begins to identify patterns and correlations, using all the relevant, accessible data. In many ways, unsupervised learning is modeled on how humans observe the world. We use intuition and experience to group things together. As we experience more and more examples of something, our ability to categorize and identify it becomes increasingly accurate. For machines, “experience” is defined by the amount of data that is input and made available. Common examples of unsupervised learning applications include facial recognition, gene sequence analysis, market research, and cybersecurity.
What is semi-supervised learning?
Semi-supervised learning is the third of four machine learning models. In a perfect world, all data would be structured and labeled before being input into a system. But since that is obviously not feasible, semi-supervised learning becomes a workable solution when vast amounts of raw, unstructured data are present. This model consists of inputting small amounts of labeled data to augment unlabeled data sets. Essentially, the labeled data acts to give a running start to the system and can considerably improve learning speed and accuracy. A semi-supervised learning algorithm instructs the machine to analyze the labeled data for correlative properties that could be applied to the unlabeled data.
As explored in depth in this MIT Press research paper, there are, however, risks associated with this model, where flaws in the labeled data get learned and replicated by the system. Companies that most successfully use semi-supervised learning ensure that best practice protocols are in place. Semi-supervised learning is used in speech and linguistic analysis, complex medical research such as protein categorization, and high-level fraud detection.
What is reinforcement learning?
Reinforcement learning is the fourth machine learning model. In supervised learning, the machine is given the answer key and learns by finding correlations among all the correct outcomes. The reinforcement learning model does not include an answer key but, rather, inputs a set of allowable actions, rules, and potential end states. When the desired goal of the algorithm is fixed or binary, machines can learn by example. But in cases where the desired outcome is mutable, the system must learn by experience and reward. In reinforcement learning models, the “reward” is numerical and is programmed into the algorithm as something the system seeks to collect.
In many ways, this model is analogous to teaching someone how to play chess. Certainly, it would be impossible to try to show them every potential move. Instead, you explain the rules and they build up their skill through practice. Rewards come in the form of not only winning the game, but also acquiring the opponent’s pieces. Applications of reinforcement learning include automated price bidding for buyers of online advertising, computer game development, and high-stakes stock market trading.
Enterprise machine learning in action
Machine learning algorithms recognize patterns and correlations, which means they are very good at analyzing their own ROI. For companies that invest in machine learning technologies, this feature allows for an almost immediate assessment of operational impact. Below is just a small sample of some of the growing areas of enterprise machine learning applications.
- Recommendation engines: From 2009 to 2017, the number of U.S. households subscribing to video streaming services rose by 450%. And a 2020 article in Forbes magazine reports a further spike in video streaming usage figures of up to 70%. Recommendation engines have applications across many retail and shopping platforms, but they are definitely coming into their own with streaming music and video services.
- Dynamic marketing: Generating leads and ushering them through the sales funnel requires the ability to gather and analyze as much customer data as possible. Modern consumers generate an enormous amount of varied and unstructured data – from chat transcripts to image uploads. The use of machine learning applications helps marketers understand this data – and use it to deliver personalized marketing content and real-time engagement with customers and leads.
- ERP and process automation: ERP databases contain broad and disparate data sets, which may include sales performance statistics, consumer reviews, market trend reports, and supply chain management records. Machine learning algorithms can be used to find correlations and patterns in such data. Those insights can then be used to inform virtually every area of the business, including optimizing the workflows of Internet of Things (IoT) devices within the network or the best ways to automate repetitive or error-prone tasks.
- Predictive maintenance: Modern supply chains and smart factories are increasingly making use of IoT devices and machines, as well as cloud connectivity across all their fleets and operations. Breakdowns and inefficiencies can result in enormous costs and disruptions. When maintenance and repair data is collected manually, it is almost impossible to predict potential problems – let alone automate processes to predict and prevent them. IoT gateway sensors can be fitted to even decades-old analog machines, delivering visibility and efficiency across the business.
Machine learning challenges
In his book Spurious Correlations, data scientist and Harvard graduate Tyler Vigan points out that “Not all correlations are indicative of an underlying causal connection.” To illustrate this, he includes a chart showing an apparently strong correlation between margarine consumption and the divorce rate in the state of Maine. Of course, this chart is intended to make a humorous point. However, on a more serious note, machine learning applications are vulnerable to both human and algorithmic bias and error. And due to their propensity to learn and adapt, errors and spurious correlations can quickly propagate and pollute outcomes across the neural network.
An additional challenge comes from machine learning models, where the algorithm and its output are so complex that they cannot be explained or understood by humans. This is called a “black box” model and it puts companies at risk when they find themselves unable to determine how and why an algorithm arrived at a particular conclusion or decision.
Fortunately, as the complexity of data sets and machine learning algorithms increases, so do the tools and resources available to manage risk. The best companies are working to eliminate error and bias by establishing robust and up-to-date AI governance guidelines and best practice protocols.
Machine learning FAQs
What’s the difference between AI and machine learning?
Machine learning is a subset of AI and cannot exist without it. AI uses and processes data to make decisions and predictions – it is the brain of a computer-based system and is the “intelligence” exhibited by machines. Machine learning algorithms within the AI, as well as other AI-powered apps, allow the system to not only process that data, but to use it to execute tasks, make predictions, learn, and get smarter, without needing any additional programming. They give the AI something goal-oriented to do with all that intelligence and data.
Can machine learning be added to an existing system?
Yes, but it should be approached as a business-wide endeavor, not just an IT upgrade. The companies that have the best results with digital transformation projects take an unflinching assessment of their existing resources and skill sets and ensure they have the right foundational systems in place before getting started.
Data science versus machine learning
Relative to machine learning, data science is a subset; it focuses on statistics and algorithms, uses regression and classification techniques, and interprets and communicates results. Machine learning focuses on programming, automation, scaling, and incorporating and warehousing results.
Data mining versus neural networks
Machine learning looks at patterns and correlations; it learns from them and optimizes itself as it goes. Data mining is used as an information source for machine learning. Data mining techniques employ complex algorithms themselves and can help to provide better organized data sets for the machine learning application to use.
Deep learning versus neural networks
The connected neurons with an artificial neural network are called nodes, which are connected and clustered in layers. When a node receives a numerical signal, it then signals other relevant neurons, which operate in parallel. Deep learning uses the neural network and is “deep” because it uses very large volumes of data and engages with multiple layers in the neural network simultaneously.
Machine learning versus statistics
Machine learning is the amalgam of several learning models, techniques, and technologies, which may include statistics. Statistics itself focuses on using data to make predictions and create models for analysis.