An A-Z of AI

Artificial Intelligence (AI) is a rapidly evolving field that can sometimes feel overwhelming due to the number of new concepts and technical terms involved. This glossary provides clear, concise definitions of key AI terms to help you better understand the technologies shaping our world. Whether you are new to AI or looking to expand your knowledge, this guide is a great resource for anyone wanting to learn more about AI and its many applications.

A/B Testing

A/B Testing is a method used to compare two different versions of something—like a website or an ad—to see which one performs better. It helps you make data-driven decisions by testing small changes to see what users prefer.

Adam Optimiser

Adam Optimiser is an algorithm used for training machine learning models. It adjusts the learning rate as training progresses, making it a popular choice for deep learning.

Adversarial AI

Adversarial AI refers to techniques used to deceive AI systems by providing misleading input, often to expose vulnerabilities in machine learning models.

AGI (Artificial General Intelligence)

AGI stands for Artificial General Intelligence, which refers to an AI system with the ability to understand, learn, and apply knowledge across a wide range of tasks, similar to a human being.

AI (Artificial Intelligence)

AI is a type of technology that enables machines to perform tasks that typically require human intelligence. This includes things like understanding speech, recognising patterns, and making decisions.

AI Agents

An AI agent is a smart digital tool that autonomously performs tasks, using data from its environment to make decisions and take actions.

AI Alignment Problem

The AI Alignment Problem is the challenge of ensuring that AI systems' goals and actions are aligned with human values and intentions, to avoid unintended consequences.

AI Automation

AI Automation involves using AI to perform tasks that were previously done manually, allowing businesses to save time and reduce costs.

AI Bias

AI Bias occurs when an AI system produces unfair or prejudiced outcomes due to biassed training data, which can lead to discrimination or incorrect results.

AI Ethics

AI Ethics is the field that focuses on the moral implications and ethical considerations of using AI technology, including fairness, transparency, and accountability.

AI Explainability

AI Explainability refers to the ability to understand and interpret how an AI system makes decisions. It is important for building trust in AI systems, especially in high-stakes applications.

AI Optimisation

AI Optimisation involves improving the efficiency and effectiveness of AI models, making them faster and more accurate at performing their tasks.

AI Planning

AI Planning is the process of creating a sequence of actions for an AI system to achieve a specific goal, often used in robotics and decision-making applications.

AI Safety

AI Safety focuses on ensuring that AI systems operate as intended without causing harm, especially as they become more advanced and autonomous.

AI-Driven Automation

AI-Driven Automation is the use of AI technologies to fully automate tasks, often leading to increased productivity and efficiency in business operations.

AI-Driven Personalisation

AI-Driven Personalisation uses AI to tailor products, services, or experiences to individual users based on their preferences and behaviour.

AI-Powered Assistants

AI-Powered Assistants are virtual assistants, like Siri or Alexa, that use AI to understand and respond to user requests, helping with tasks such as scheduling or answering questions.

Artificial General Intelligence

Artificial General Intelligence (AGI) is an advanced form of AI that can understand, learn, and apply knowledge across different domains, similar to human intelligence.

Artificial Intelligence Alignment Problem

The Artificial Intelligence Alignment Problem involves ensuring that AI systems act in accordance with human values and do not produce unintended or harmful outcomes.

Artificial Intelligence Automation

Artificial Intelligence Automation is the use of AI to perform tasks automatically, often replacing manual labour and improving efficiency.

Artificial Intelligence Bias

Artificial Intelligence Bias occurs when AI systems make decisions that are unfair or biased due to problems in the data they were trained on.

Artificial Intelligence Ethics

Artificial Intelligence Ethics involves the study of ethical issues related to the development and use of AI, including questions of fairness, accountability, and societal impact.

Artificial Intelligence Explainability

Artificial Intelligence Explainability refers to how transparent and understandable an AI system's decision-making process is, which helps users trust the system.

Artificial Intelligence Optimisation

Artificial Intelligence Optimisation is about making AI systems more efficient, accurate, and effective at solving specific tasks.

Artificial Intelligence Planning

Artificial Intelligence Planning is a process used by AI to determine a set of actions to achieve a specific goal, often used in robotics and automation.

Artificial Intelligence Safety

Artificial Intelligence Safety ensures that AI technologies are developed and used in a way that minimises risks and prevents unintended harmful consequences.

Artificial Intelligence-Driven Automation

Artificial Intelligence-Driven Automation refers to the use of AI to perform repetitive or complex tasks automatically, freeing up human workers for more creative or strategic activities.

Artificial Intelligence-Driven Personalisation

Artificial Intelligence-Driven Personalisation tailors content, recommendations, and services to individual users by analysing their behaviour and preferences.

Artificial Intelligence-Powered Assistants

Artificial Intelligence-Powered Assistants are AI-based tools, such as chatbots or virtual assistants, designed to help users by answering questions or performing tasks.

Autoencoders

Autoencoders are a type of neural network used for unsupervised learning. They are often used to compress data, like reducing the dimensions of an image while keeping its important features.

Bagging and Boosting

Bagging and Boosting are machine learning techniques used to improve the accuracy of models by combining multiple weaker models to create a stronger overall model.

Batch Normalisation

Batch Normalisation is a technique used in training neural networks to make the training process faster and more stable by normalising the input data.

Bayesian Inference

Bayesian Inference is a method of statistical inference that uses Bayes' theorem to update the probability for a hypothesis as more evidence becomes available.

Bayesian Networks

Bayesian Networks are graphical models that represent the probabilistic relationships between a set of variables, often used for decision-making and reasoning under uncertainty.

Bias in AI

Bias in AI occurs when an AI model produces biased or unfair results, often due to biased data used during training. This can lead to unequal treatment or inaccurate predictions.

Bias in Machine Learning

Bias in Machine Learning refers to errors or unfairness that arise in machine learning models due to problems with the training data or how the model was built.

Bias-Variance Tradeoff

The Bias-Variance Tradeoff is a concept in machine learning that describes the balance between the model's complexity and its accuracy. A good balance helps the model generalise well to new data.

Chatbots

Chatbots are AI programs designed to simulate conversation with human users, often used for customer support or answering common questions on websites.

Classification

Classification is a type of machine learning task where the goal is to assign labels to input data. For example, classifying emails as either "spam" or "not spam".

Clustering

Clustering is an unsupervised learning technique that groups similar data points together, often used for market segmentation or identifying patterns in data.

CNN (Convolutional Neural Networks)

Convolutional Neural Networks (CNNs) are a type of neural network commonly used in image recognition and processing. They are effective at detecting features in visual data.

Cognitive Computing

Cognitive Computing refers to technologies that try to simulate human thought processes, helping computers understand and respond to complex information in a human-like way.

Collaborative Filtering

Collaborative Filtering is a technique used in recommendation systems to predict what users might like based on their past behaviour and the preferences of others.

Computer Vision

Computer Vision is a field of AI that enables machines to interpret and understand visual information from the world, such as recognizing objects in photos or videos.

Conversational Agents

Conversational Agents are AI systems designed to communicate with users, usually through text or voice, to help answer questions or complete tasks.

Convolutional Neural Networks

Convolutional Neural Networks (CNNs) are a type of deep learning model used primarily for analysing visual data, such as recognizing objects in images.

Cross-Validation

Cross-Validation is a technique used to assess the performance of a machine learning model by dividing the data into multiple subsets to test the model's reliability.

Data Mining in AI

Data Mining in AI refers to the process of extracting useful information and patterns from large datasets, often used to make predictions or support decision-making.

Decision Boundary

A Decision Boundary is a line or surface that separates different classes in a classification problem, helping the model decide which class a new data point belongs to.

Decision Trees

Decision Trees are a type of model used for both classification and regression tasks. They work by splitting the data into branches based on feature values, ultimately leading to a decision.

Deep Learning

Deep Learning is a subset of machine learning that uses neural networks with many layers to analyse and learn from data. It is commonly used for image recognition, natural language processing, and more.

Dimensionality Reduction

Dimensionality Reduction is the process of reducing the number of features in a dataset while keeping important information, making the data easier to analyse.

Dropout

Dropout is a regularisation technique used in neural networks to prevent overfitting by randomly dropping units (neurons) during training.

Early Stopping

Early Stopping is a technique used during model training to prevent overfitting by stopping the training process once the model's performance stops improving on a validation dataset.

End-to-End Learning in AI

End-to-End Learning in AI refers to training an AI system to perform a complete task from start to finish, without breaking it into smaller sub-tasks, often used in complex systems like self-driving cars.

Ensemble Learning

Ensemble Learning is a technique where multiple machine learning models are combined to create a more powerful and accurate model than any of the individual models alone.

Epochs

Epochs are the number of times a machine learning model goes through the entire training dataset during the training process.

Ethical AI

Ethical AI is the practice of designing and using AI technologies in a way that is responsible and respects human rights, ensuring fairness and transparency.

Ethics in AI

Ethics in AI refers to the study and practice of ensuring that AI technologies are developed and used in ways that are ethical, fair, and beneficial to society

Expert Systems

Expert Systems are AI programs that use a database of knowledge and a set of rules to provide advice or solve specific problems, often used in medical diagnosis or technical support.

Explainable AI

Explainable AI refers to AI systems that provide clear and understandable explanations of their decisions, making it easier for humans to trust and verify the outcomes.

Exploration vs. Exploitation

Exploration vs. Exploitation is a concept in reinforcement learning that involves deciding whether to try new actions (exploration) or stick with known actions that give the best rewards (exploitation).

Facial Recognition

Facial Recognition is a technology that uses AI to identify and verify people by analysing their facial features, often used for security or authentication.

Federated Learning

Federated Learning is a type of machine learning where the model is trained across multiple devices without sharing the data, helping to protect user privacy.

Feature Engineering

Feature Engineering is the process of selecting, modifying, or creating features (input variables) that help a machine learning model perform better.

Few-Shot Learning

Few-Shot Learning is a type of machine learning where the model learns to recognize new categories with only a few examples, making it more efficient in situations with limited data.

Fuzzy Logic in AI

Fuzzy Logic in AI is a form of reasoning that allows for uncertain or approximate values, similar to human decision-making, often used in control systems and consumer electronics.

GANs (Generative Adversarial Networks)

Generative Adversarial Networks (GANs) are a type of deep learning model that consists of two neural networks working against each other to generate realistic data, such as images or music.

Generalisation in AI

Generalisation in AI refers to a model's ability to perform well on new, unseen data, indicating that it has learned the underlying patterns rather than just memorising the training data.

Generative Adversarial Networks

Generative Adversarial Networks (GANs) are deep learning models consisting of two parts—a generator and a discriminator—that work together to generate realistic data, such as images or videos.

Generative AI

Generative AI refers to AI models that create new content, such as text, images, or music, by learning from existing examples.

Gradient Boosting

Gradient Boosting is a machine learning technique that builds an ensemble of weak models, usually decision trees, to create a strong predictive model.

Gradient Descent

Gradient Descent is an optimisation algorithm used in machine learning to minimise the error of a model by iteratively adjusting its parameters in the direction that reduces the error.

Grid Search

Grid Search is a technique used in machine learning to find the best hyperparameters for a model by systematically trying different combinations.

Hierarchical Clustering

Hierarchical Clustering is a type of clustering technique that groups similar data points into clusters in a tree-like structure, often used for data analysis.

Human-AI Collaboration

Human-AI Collaboration refers to situations where humans and AI systems work together to complete tasks, combining human intuition and creativity with AI's speed and precision.

Hybrid AI Systems

Hybrid AI Systems combine different types of AI methods, such as symbolic AI and machine learning, to take advantage of the strengths of each approach.

Hyperparameter Tuning

Hyperparameter Tuning is the process of optimising the hyperparameters (settings) of a machine learning model to improve its performance.

K-Means Clustering

K-Means Clustering is an unsupervised learning algorithm that groups data into clusters based on similarity, often used for market segmentation and pattern recognition.

K-Nearest Neighbors (KNN)

K-Nearest Neighbors (KNN) is a simple machine learning algorithm used for classification and regression tasks. It assigns labels based on the majority vote of the nearest data points.

Knowledge Graphs in AI

Knowledge Graphs in AI are structures that represent relationships between entities, helping AI systems understand and reason about the connections between different pieces of information.

Knowledge Representation

Knowledge Representation refers to how information is structured and stored so that an AI system can use it effectively, often using graphs, logic, or rules.

Learning Rate

Learning Rate is a hyperparameter that controls how much the model's parameters are adjusted during training. A high learning rate can lead to faster training, while a low learning rate can provide more precise updates.

Linear Regression

Linear Regression is a basic machine learning technique used to predict the value of a variable based on the relationship with another variable. It assumes a straight-line relationship between variables.

Logistic Regression

Logistic Regression is a classification algorithm used to predict binary outcomes (e.g., yes/no, true/false) based on input features.

LSTM (Long Short-Term Memory)

Long Short-Term Memory (LSTM) is a type of recurrent neural network (RNN) used for tasks that involve sequences, such as language modelling and time series forecasting. It is designed to remember information over long periods.

Machine Learning

Machine Learning is a branch of AI that involves teaching computers to learn from data without being explicitly programmed. It allows systems to make predictions or decisions based on patterns found in the data.

Markov Decision Process

A Markov Decision Process (MDP) is a mathematical framework used in reinforcement learning to model decision-making situations, where outcomes depend on both current actions and random factors.

Meta-Learning

Meta-Learning, also known as "learning to learn," is a subfield of machine learning where the model learns to adapt to new tasks quickly, often by understanding how different tasks are related.

Monte Carlo Methods

Monte Carlo Methods are statistical techniques used to solve problems by running simulations with random sampling, often used in AI for decision-making under uncertainty.

Multi-Agent Systems

Multi-Agent Systems are systems that involve multiple interacting agents, each with their own goals, that work together or compete, often used in simulations and game theory.

Natural Language Processing (NLP)

Natural Language Processing (NLP) is a branch of AI that focuses on helping computers understand, interpret, and generate human language, used in applications like chatbots and language translation.

Neural Networks

Neural Networks are a type of machine learning model inspired by the human brain, consisting of layers of interconnected nodes (neurons) that learn to recognize patterns in data.

Neuro-Symbolic AI

Neuro-Symbolic AI combines neural networks with symbolic reasoning to create AI systems that can learn from data while also understanding and applying logical rules.

Overfitting

Overfitting is a problem in machine learning where a model learns the training data too well, including the noise, resulting in poor performance on new, unseen data.

Pattern Recognition in AI

Pattern Recognition in AI involves identifying patterns and regularities in data, such as recognizing faces in a photo or detecting fraud in transactions.

PCA (Principal Component Analysis)

Principal Component Analysis (PCA) is a dimensionality reduction technique that reduces the number of features in a dataset while keeping the most important information, making it easier to visualise and analyse.

Perceptron

A Perceptron is a simple type of neural network used for binary classification tasks. It is the building block of more complex neural networks.

Predictive Analytics

Predictive Analytics is the use of data, statistical algorithms, and AI to make predictions about future outcomes, often used in business to forecast sales or identify trends.

Principal Component Analysis

Principal Component Analysis (PCA) is a dimensionality reduction method that transforms a large set of variables into a smaller one, retaining the most important information to simplify analysis.

Random Forests

Random Forests are an ensemble learning technique that uses multiple decision trees to make more accurate and reliable predictions, often used for classification and regression tasks.

Random Search

Random Search is a method for finding the best hyperparameters for a machine learning model by randomly trying different combinations.

Recurrent Neural Networks

Recurrent Neural Networks (RNNs) are a type of neural network used for tasks that involve sequences, such as time series analysis or language modelling. They have connections that allow them to maintain information over time.

Regression

Regression is a type of machine learning used to predict continuous values, such as predicting house prices based on size and location.

Regularisation

Regularisation is a technique used in machine learning to prevent overfitting by adding a penalty to the model's complexity, encouraging it to learn simpler patterns.

Reinforcement Learning

Reinforcement Learning is a type of machine learning where an agent learns to make decisions by interacting with an environment and receiving rewards or penalties based on its actions.

Residual Neural Networks (ResNet)

Residual Neural Networks (ResNet) are a type of deep learning model that include shortcut connections to help train very deep networks more effectively, often used in computer vision tasks.

Retrieval-Augmented Generation (RAG)

Retrieval-Augmented Generation (RAG) is a type of generative AI model that combines information retrieval with text generation to produce more informative and accurate outputs.

Robotics in AI

Robotics in AI involves the use of AI technologies to enable robots to perform tasks autonomously, including perception, decision-making, and control.

Self-Supervised Learning in AI

Self-Supervised Learning in AI is a type of learning where the system generates its own labels for training, often used in situations where labelled data is scarce.

Sentiment Analysis

Sentiment Analysis is a natural language processing task that involves determining the sentiment or emotional tone behind a piece of text, such as identifying whether a customer review is positive or negative.

SGD (Stochastic Gradient Descent)

Stochastic Gradient Descent (SGD) is an optimisation algorithm used to train machine learning models by updating the parameters based on a subset of the training data, making it faster and more efficient.

Speech Recognition

Speech Recognition is a technology that uses AI to convert spoken language into text, enabling voice commands and transcription services.

Speech Synthesis Technology

Speech Synthesis Technology, also known as text-to-speech, uses AI to convert written text into spoken words, often used in virtual assistants and accessibility tools.

Supervised Learning

Supervised Learning is a type of machine learning where the model is trained using labelled data, meaning the desired output is known, allowing the model to learn to predict new outputs.

Support Vector Machines (SVM)

Support Vector Machines (SVM) are a type of machine learning algorithm used for classification and regression tasks. They work by finding the optimal boundary that separates different classes of data.

Symbolic AI

Symbolic AI is a type of AI that represents knowledge using symbols and logical rules, focusing on reasoning and problem-solving, rather than learning from data.

Technological Singularity

The Technological Singularity is a hypothetical future point where technological growth becomes uncontrollable and irreversible, leading to unforeseen changes in society, often due to advanced AI.

Transfer Learning

Transfer Learning is a machine learning technique where a model trained on one task is adapted to perform a different but related task, saving time and improving performance.

Turing Test

The Turing Test is a test of a machine's ability to exhibit human-like intelligence. If a machine can engage in conversation indistinguishably from a human, it is said to have passed the test.

Underfitting

Underfitting occurs when a machine learning model is too simple to capture the underlying patterns in the data, resulting in poor performance on both training and new data.

Unsupervised Learning

Unsupervised Learning is a type of machine learning where the model learns from unlabeled data, meaning it must find patterns and relationships in the data without guidance.

Zero-Shot Learning

Zero-Shot Learning is a type of machine learning where the model can recognize new classes that it has not seen before, based on knowledge of related classes.

‍