Following the path of digitalization in Slovenia and Europe: Can machines learn more than humans?

Machine learning, which is considered a subset of artificial intelligence, is not a new concept. The term itself originated in the 1950s and was coined by IBM’s Arthur Samuel, based on his research into computers playing checkers (also called draughts, it is a thinking game played on a chessboard where players move figures or tiles diagonally). In the competition between the computer and the American master of checkers, the computer won, which was astonishing, and the result opened up an unlimited world of possibilities.

Today, of course, machine learning has gone far beyond simple computer games, and we have been living with it for quite some time, knowingly or unknowingly. It is used, for example, in financial services, public administration, healthcare, retail, transport and energy industries; in short, in most industries dealing with large amounts of data. In the following, we will take a closer look at what machine learning is, how it works and where it is already being used. So, let’s look at how “trained” machines already are.

What is machine learning?

Machine learning studies computers using algorithms to improve through experience and data automatically. Machine learning algorithms build a model based on sample data, known as training data so that they can predict or make decisions without being explicitly programmed to do so.

English mathematician, computer scientist and cryptologist Alan Turing, in his seminal article in 1950, introduced a reference standard for proving machine intelligence, according to which a machine must be intelligent and responsive in a way indistinguishable from a human. He wrote: “Machine Learning is an application of artificial intelligence where a computer/machine learns from past experiences (input data) and makes future predictions. The performance of such a system should be at least human level. “

A more technical definition was given in 1997 by American computer scientist and professor Tom M. Mitchell: “A computer program is said to learn from experience E concerning some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.”

Example: A handwriting recognition learning problem:

• Task T: recognising and classifying handwritten words within images
• Performance measure P: per cent of words correctly classified, accuracy
• Training experience E: a dataset of handwritten words with given classifications

To perform task T, the system learns from the dataset provided. A dataset is a collection of many examples. An example is a collection of features.

How does it work?

The American multinational developer of analytics software, SAS Institute, states on its website that machine learning is generally divided into four types: supervised learning, unsupervised learning, semi-supervised learning and reinforcement learning.

Supervised learning: algorithms are trained using labelled examples, such as an input where the desired output is known. The learning algorithm receives a set of inputs and the corresponding correct outputs, and the algorithm learns by comparing its actual output with the right outputs to find errors. It then modifies the model accordingly.

Supervised learning is commonly used in applications where historical data predicts likely future events. For example, the algorithm can anticipate when credit card transactions are likely to be fraudulent or which insurance customer is expected to make a claim.

Unsupervised learning: is used against data that has no historical labels. The system is not told the “right answer.” The algorithm must figure out what is being shown. The goal is to explore the data and find some structure within it.

Unsupervised learning works well on transactional data. It can, for example, identify segments of customers with similar attributes who can then be treated similarly in marketing campaigns. These algorithms are also used to recommend items and identify data outliers.

Semi-supervised learning: This type of machine learning uses a certain amount of labelled data and a large amount of unlabelled data to train the model. The aim of this is to classify some unlabelled data using labelled data.

Enhanced learning: often used for robotics, gaming, and navigation. The algorithm discovers which actions yield the most significant rewards through trial and error with reinforcement learning. This type of learning has three primary components: the agent (the learner or decision-maker), the environment (everything the agent interacts with) and actions (what the agent can do). The goal is for the agent to choose actions that maximise the expected reward over a given amount of time.

Structured data and unstructured data are used in machine learning.  A structured data type is predefined, labelled, and well-formatted before being stored in a data storage. Unstructured data is in native format, and it’s not processed until it is used.

Abhay Parashar, a computer scientist and software developer, explains on the KD Nuggets website how the development and production of a machine learning model are carried out:

• Data Collection: The first stage of any kind of machine learning model. In this stage, the appropriate data is decided, and then it is collected using some algorithm or manually.
• Data Processing: In this stage, the data collected in the first stage is pre-processed by handling all the null values, categorical data, etc. Also, in the same stage, the data features are made in the same range if they are not already.
• Model Building: In this stage, first, we choose appropriate algorithms to create the model, and then with the use of a specific programme, the model is built.
• Model Evaluation: After the model is created, it is evaluated using some techniques of statistics like accuracy score, z score, accuracy matrix, and more.
• Model Saving and Testing: After a successful evaluation of the model, it is saved for future use, and real-time testing is done using it.

Examples of the machine learning used in everyday life?

Banks and other businesses in the financial industry use machine learning technology mainly in two areas: to identify important insights in data and to prevent fraud. The insights can identify investment opportunities or help investors know when to trade. They can also use data mining to recognise clients with high-risk profiles or use cyber-surveillance to pinpoint warning signs of fraud.

Government agencies such as public safety and utilities have a particular need for machine learning. According to the SAS Institute, they have multiple data sources that can be mined for insights. Data analysis can identify ways to increase efficiency and save money. Machine learning can also help detect fraud and minimize entity theft.

In the health care industry, machine learning can help medical experts analyse data to identify trends or red flags that may lead to improved diagnoses and treatment. Retailers rely on machine learning to capture data, analyse it and use it to personalise a shopping experience, implement a marketing campaign, price optimisation, merchandise planning, and customer insights. Machine learning is partly to blame for the fact that annoying ads are displayed everywhere online.

In the energy industries, it is used to analyse minerals in the ground, predict refinery sensor failure, streamline oil distribution to make it more efficient and cost-effective, etc. The data analysis and modelling aspects of machine learning are essential tools to delivery companies, public transportation, and other transportation organisations.

Machine learning technology can offer us a lot, as evidenced by its many applications. The European Union is also aware of this, as it has included machine learning, alongside artificial intelligence, in its digital strategy. These technologies are closely intertwined, along with others that we have already written about. However, we forgot to mention that there is no concern about the machines conspiring against people, as humans create the model and set the framework for it.

Author: Rok Žontar

Keywords: machine learning, AI, technology, examples, use.

Disclaimer:

This article is part of joint project of the Wilfried Martens Centre for European Studies and the Anton Korošec Institute (INAK) Following the path of digitalization in Slovenia and Europe. This project receives funding from the European Parliament. 

The information and views set out in this article are those of the author and do not necessarily reflect the official opinion of the European Union institutions/Wilfried Martens Centre for European Studies/ Anton Korošec Institute. Organizations mentioned above assume no responsibility for facts or opinions expressed in this article or any subsequent use of the information contained therein.