Artificial Intelligence
Training Data
Plain English
The glossaryDataData is raw, uninterpreted information collected and stored so it can be analysed, processed, or used to inform decisions. On its own it carries no meaning; context and interpretation are what make it useful.Open glossary term used to teach AI.
Definition
guideTraining DataWhat training data is, how it shapes what an AI model knows and assumes, and what product and design teams need to understand about its role in quality.Open guide is the glossaryDatasetA dataset is a structured collection of data used for analysis, training models, or processing.Open glossary term used to teach a glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term how to perform a task.
In practice
Used to train glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term by exposing them to examples, enabling them to learn glossaryPatternA pattern is a reusable solution to a design problem that recurs across products and contexts. It captures an approach already shown to work, so a team solves the problem once rather than every time it appears.Open glossary term and relationships.
In context
It determines the ceiling. Every limitation and every glossaryBiasBias is a systematic distortion in thinking or data that affects the accuracy of research or decision-making.Open glossary term in it will appear in the glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term's glossaryBehaviourBehaviour refers to how users interact with a system, including actions, patterns, and responses.Open glossary term, usually in the cases you did not think to test.
The reality
Poor or biased guideTraining DataWhat training data is, how it shapes what an AI model knows and assumes, and what product and design teams need to understand about its role in quality.Open guide glossaryLeadA lead is a potential customer who has shown interest in a product or service, typically by providing contact information or engaging with content.Open glossary term to poor or biased glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term performance.
Compared with
Training Data vs Test Data
guideTraining DataWhat training data is, how it shapes what an AI model knows and assumes, and what product and design teams need to understand about its role in quality.Open guide teaches the glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term. Test glossaryDataData is raw, uninterpreted information collected and stored so it can be analysed, processed, or used to inform decisions. On its own it carries no meaning; context and interpretation are what make it useful.Open glossary term is held back to check what it learned generalises. Letting the two overlap produces excellent scores and a model that fails in production.
FAQ
Common questions
A few practical answers to the questions that usually come up around this term.
What is training data?
It is glossaryDataData is raw, uninterpreted information collected and stored so it can be analysed, processed, or used to inform decisions. On its own it carries no meaning; context and interpretation are what make it useful.Open glossary term used to train glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term.
Why is training data important?
It determines how well a glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term performs.
What happens with bad training data?
glossaryModelA model is a system or representation used to process data and generate outputs, often trained to perform specific tasks.Open glossary term produce inaccurate or biased outputs.
How is training data created?
By collecting and glossaryLabellingLabelling is the practice of naming content, categories, and interface elements in a way that is clear and meaningful to users. It directly affects how users understand and navigate a product.Open glossary term relevant examples.
Related Services
Related Guides
Related Terms
