Machine learning — an application of artificial intelligence (AI) that allows systems to learn and improve from experience without being explicitly programmed. 

In 1952, the term “machine learning” was first used. Almost 70 years later, we are using machine learning on a daily basis: Face and Touch ID, Siri, Alexa, Uber, and even Computer-Aided Detection. For example, CAD software can spot 52% of breast cancer cells, a year before patients are diagnosed. What is machine learning technology, and what are its application areas? Let’s get all these straightened out.

How did it all begin?

The 50s

  • Machine learning gets a definition.
  • Alan Turing creates the “Turing Test” to determine if a computer has real intelligence. A computer must be able to fool a human into believing it is also human.
  • Arthur Samuel writes the first computer learning program (the game of checkers). An IBM computer improves at the game the more it plays.
  • Frank Rosenblatt designs the first neural network for computers (the perceptron) to simulate the thought processes of the human brain.

The 60s

  • Computers begin to use a basic pattern recognition due to the “nearest neighbor” algorithm. It is used to map a route for traveling salesmen.

The 70s

  • Students at Stanford University invent the “Stanford Cart”. This remotely controlled, TV-equipped mobile robot can navigate obstacles in a room on its own.

The 80s

  • Gerald Dejong introduces Explanation Based Learning (EBL). According to this concept, a computer analyses training data and creates a general rule that it can follow by discarding unimportant data.
  • Terry Sejnowski invents NetTalk, an artificial neural network that learns to pronounce words the same way a baby does.

The 90s

  • The onset of the data-driven approach. Scientists begin creating programs for computers to analyze large amounts of data and learn from the results.
  • IBM’s Deep Blue beats the world champion at chess.

The 2000s

  • Geoffrey Hinton introduces the term “deep learning”. It is used to explain new algorithms that let computers “see” and distinguish objects and text in images and videos.
  • The Microsoft Kinect tracks 20 human features, allowing humans to interact with a computer via movements and gestures.
  • IBM’s Watson beats human competitors at Jeopardy (a classic game show).
  • Google Brain, a deep learning artificial intelligence research team at Google is formed. It combines open-ended machine learning research with systems engineering and Google-scale computing resources.
  • Google’s X Lab launches a machine learning algorithm that is able to autonomously browse YouTube videos to identify the videos that contain cats.
  • Facebook develops DeepFace. This is a software algorithm that can recognize or verify individuals on photos.
  • Amazon launches its own machine learning platform.
  • Microsoft creates the Distributed Machine Learning Toolkit, which enables the efficient distribution of machine learning problems across multiple computers.
  • Google’s AI algorithm beats a professional player at the Chinese board game Go.
  • Google AI  presents BERT (Bidirectional Encoder Representations from Transformers), a new bidirectional language model that will result in more accurate machine translation, chatbot behavior, automated email responses, and customer review analysis.

How does machine learning work?

There are two types of techniques employed by machine learning: supervised and unsupervised learning.

Supervised learning trains a model on known input and output data so that it can collect data or produce a data output from the previous experience. It helps you to solve various types of real-world computation problems. Supervised learning uses classification and regression techniques to develop predictive models.

Classification techniques classify input data into categories, for example, whether an email is genuine or spam, or whether a tumor is cancerous or benign. Typical applications include medical imaging, speech recognition, and credit scoring.

Regression techniques are used while working with a data range or if the nature of your response is a real number, such as temperature or the time until failure for a piece of equipment. These models predict continuous responses, for example, changes in temperature or fluctuations in power demand.

Unsupervised learning finds all kinds of unknown patterns in data. It helps to find features that can be useful for categorization. The most common unsupervised learning technique is clustering. Applications for cluster analysis include gene sequence analysis, market research, and object recognition. 

What is machine learning used for?

In industries

Financial services. Financial institutions and businesses use machine learning for two purposes: to identify important insights (e.g., investment opportunities, clients with high-risk profiles) and prevent fraud.

Government. Government agencies use machine learning to identify ways to increase efficiency, save money, detect fraud, and minimize identity theft.

Healthcare. The technology helps medical experts analyze data to identify trends or red flags that may lead to improved diagnosis and treatment. 

Retail. Websites recommend items you might want to buy based on previous purchases using machine learning to analyze your buying history. 

Oil and gas. Machine learning helps to find new energy sources, analyze minerals in the ground, predict refinery sensor failure, streamline oil distribution to make it more cost-effective. 

Transportation. Data analysis and modeling are important tools for delivery companies and public transportation to make routes more efficient and predict potential problems.

In everyday life

Virtual personal assistants. Some of the popular examples are Siri, Alexa, and Google Now. Machine learning is an important part of these PAs as they collect and process the information on the basis of your previous involvement with them. 

Traffic predictions. GPS navigation services save our current location at a central server for managing traffic. This data is then used to build a map of the current traffic. 

Online transportation. When booking a cab, the app defines price surge hours by predicting the rider demand.  

Video surveillance. Such systems are powered with AI that tracks unusual behavior of people like standing motionless for a long time, stumbling, or napping on benches, etc. And when such activities are reported, machine learning collects them and uses further, helping to improve the surveillance services. 

Social media. Here are a few examples of machine learning in action: face recognition and people you may know on Facebook, similar pins on Pinterest.

Online customer support. Websites use chatbots that extract information from the website and present it to the customers.

Search engine result refining. Search engines use machine learning to make your search results more accurate. For example, if you reach the second or third page of the search results and don’t open any of them, the search engine estimates that the results served did not match requirements. 

On a final note

Why is machine learning important? — It helps us to deal with large amounts of data and improve our experience in almost every area of life.  What’s more, machine learning is the drive to labor automation, which allows businesses to profit from the more human and creative side of work.