AI Glossary: A beginner’s guide to the definitions of Artificial Intelligence
Trying to get your head around what Artificial Intelligence (AI) is and what it might mean for the future could be extremely confusing. Below we guide you through the acronyms, jargon and key definitions associated with the technology to improve your understanding of the terms closely linked with AI.
Algorithms are a form of automated instruction that is used for calculation, data processing, and reasoning. The algorithms are what turn a data set into a model – processing the data in such a way that the machine can read it and apply the rules moving forward. A set of algorithms is what makes up the basis of Artificial Intelligence.
Machine learning is a core branch of AI. It focuses on building applications that learn from data and can therefore improve their accuracy over time, without being explicitly programmed to do so. In this sense we can ‘teach’ computers to think the way a human would when carrying out tasks.
For example, when we bump into an object, we take extra care to learn where it was to avoid doing so again in the future. A typical computer in a machine can’t do this. That’s why, without programming, your Roomba would continue to hit the same objects as it cleaned your floor.
Machine learning uses algorithms that are ‘trained’ to find patterns in massive amounts of data in order to make decisions and predictions based on new data. The examples of this in action are already all around us. Voice assistants learn the way we speak to improve what data they search for when looking up our requests, spam detectors stop unwanted emails reaching our primary inbox, and the first self-driving cars are hitting the roads.
Deep learning is a subset of machine learning (all deep learning is machine learning, but not all machine learning is deep learning). Deep learning is different from classical machine learning due to the types of data that it works with and the methods in which it learns.
The key difference is that while machine learning algorithms leverage structured, labelled data to make predictions, with deep learning some of data pre-processing that is typically involved with machine learning is eliminated.
Deep learning models therefore require large amounts of data that passes through multiple layers of calculations, applying weights and biases in each successive layer to continually adjust and improve results.
For instance, let’s say that we had a set of photos of different animals, and we wanted to categorise them by their name e.g. “dolphin”, “dog”, “elephant”. Deep learning algorithms can determine which features are most important to distinguish each animal from another. In machine learning, this would be established manually by a human expert who would tell the algorithm to look for differences in ears, for example.
Big data refers to the analysis of data sets that are too large or complex to be dealt with by traditional data-processing application software.
Characteristics of big data include high volume, high velocity and high variety.
Deep learning requires big data because only this type of data sets can isolate hidden patterns and to find answers to questions we otherwise couldn’t. As such, big data makes it possible for machine learning applications to do what they were built to do: learn and acquire skills. Without big data it wouldn’t be possible to develop and train the intelligent algorithms and predictive models that make AI a game-changing technology.
Got any other terms you’d like to see added to this list? Let us know in the comments below.
You can also read our other blogs about the applications of AI here: