MACHINE LEARNING

Increasing amounts of data

Our world is fully connected today. In almost every area of life – whether private or professional – data is being generated: we accept cookies while browsing the internet, pay by credit card, use apps and social media, or rely on sensors in machines, vehicles, and everyday devices. Weather stations and IoT systems also continuously provide new streams of information.

Machine learning: systems that learn from data

Companies have been using business intelligence software for years to analyze data and generate reports. However, these data sets can unlock even greater potential when combined with Machine Learning. It is a subset of artificial intelligence and describes the ability of systems to generate knowledge from experience – in other words, from data. Software learns rules from examples, which it can then generalize and apply to new situations.

For example, a Machine Learning model can learn to distinguish apples from pears. By analyzing training data – in this case, photos of apples and pears – the system identifies patterns and regularities. Based on this learning, the model or algorithm can then classify whether a new, previously unseen image shows an apple or a pear. The result is validated, and the model is improved over time.

The quality of a Machine Learning model is measured by its error rate. The principle is simple: the more relevant data a model has access to – essentially, the more “practice” it gets – the better it learns and the more accurate its predictions become.

Artificial Intelligence

Machine Learning

Supervised and Unsupervised Learning

Supervised vs unsupervised Learning

 

In this example, machine learning takes place using verifiable data as supervised learning: the test data is given a correct label (apple or pear) by humans in advance. The algorithm can then compare its prognoses with the actual results and improve itself step by step.

This is different from machine learning using the unsupervised learning principle: no learning target is provided. The machine learning model is fed with data without any prior definition of what this data means. The machine learning algorithm recognises patterns and divides data up into clusters or categories without knowing what categories these are, i.e. under which labels they fall (such as ‘apple’ and ‘pear’).

The majority of machine learning processes used in practice entail supervised learning.

Supervised vs unsupervised Learning-Struktur

 

Deep Learning - what is that?

Deep Learning is a subset of Machine Learning and is based on artificial neural networks. Through training methods, the system develops artificial intelligence capable of learning from data. Information is continuously processed, and new inputs are integrated with what has already been learned, allowing the model to improve over time. Based on these learning processes, the system can generate predictions and make decisions, which are continuously reviewed and adjusted. Human intervention is generally required only initially. Deep Learning enables systems to learn autonomously and is particularly suitable for applications with large datasets, from which patterns and models can be derived. The foundation consists of artificial neural networks, which are repeatedly reconnected and optimized during the learning process.

Machine learning in practice

A whole range of technology that we use daily in our private lives is already based on machine learning algorithms. Examples are facial recognition in photos and videos, speech recognition on mobile phones, translation programmes online and computer spam filters.