The Ad Agency AI Dictionary

Data Augmentation

Techniques that create new training samples by transforming existing data, improving model robustness when data is scarce.

Data Cleansing

The process of detecting and correcting errors or inconsistencies in data to improve quality and reliability.

Data Distribution

The statistical properties of how values are spread across a dataset; understanding distribution helps in modelling and targeting.

Data Drift

When the data used by a model changes over time, potentially degrading performance; monitoring drift helps keep campaigns accurate.

Data Governance

Policies and processes that ensure data is managed, secured and used ethically across an organisation.

Data Integration

Combining data from multiple sources into a unified view to enable comprehensive analysis and activation.

Data Lake

A storage repository that holds raw data in its native format until needed for analytics or modelling.

Data Labeling

Assigning meaningful labels to data, often using human annotators or AI, to train supervised learning models.

Data Leakage

Accidental inclusion of information from outside the training set that gives a model unrealistic performance during training.

Data Lineage

Documentation of where data originates, how it moves and how it is transformed across systems.

Data Mining

Discovering patterns and relationships in large datasets through algorithms and statistical techniques.

Data Normalization

Rescaling numeric data to a common range to improve model convergence and interpretation.

Data Pipeline

The sequence of processes that extract, transform and load data from sources into storage or analytical systems.

Data Preprocessing

Preparing raw data for analysis through cleaning, normalization and transformation.

Data Privacy

Policies and techniques that ensure personal data is collected, stored and used in compliance with regulations.

Data Quality

The reliability, accuracy and completeness of data; high‑quality data underpins trustworthy marketing insights.

Data Science

The field combining statistics, computer science and domain knowledge to extract insights from data.

Data Streaming

Processing data continuously as it arrives, enabling real‑time analytics and decision‑making.

Data Transformation

Converting data from one format or structure to another, such as aggregating events into daily metrics.

Data Visualization

Presenting data in charts or dashboards to make patterns and trends understandable at a glance.

Data Warehouse

A centralized repository that stores structured data from various sources for analysis and reporting.

Deep Learning

A subset of machine learning that uses multi‑layer neural networks to learn complex patterns in data.

Decision Boundary

The surface separating different classes in a model’s input space; understanding boundaries helps evaluate model behaviour.

Decision Tree

A model that makes sequential decisions by splitting data based on feature values; interpretable and useful for marketing segmentation.

Defensive AI

Techniques designed to protect systems against adversarial attacks and ensure model robustness.

Demographic Parity

A fairness criterion requiring that outcomes are independent of sensitive attributes like age or gender.

Differential Privacy

A framework for analysing data while protecting individual privacy by adding controlled noise to results.

Dimensionality Reduction

Reducing the number of variables in a dataset to simplify analysis and visualization.

Distributed Computing

Using multiple computers to process data or train models more quickly and handle larger workloads.

Domain Adaptation

Adjusting models trained on one domain so they perform well on a related, but different, domain.

Domain Randomization

Simulating varied environments to train models that generalize better to real‑world conditions.

Dropout

A regularization technique that randomly disables neurons during training to prevent overfitting.

Dynamic Creative Optimization or DCO

Using AI to assemble personalized ads in real time by mixing headlines, images and calls‑to‑action based on user data:

Dynamic Programming

A method for solving complex problems by breaking them into simpler subproblems, used in optimisation.

Dynamic Routing

Mechanism used in capsule networks to determine how much lower‑level capsules contribute to higher‑level ones.

Dynamic Time Warping

A technique for aligning sequences of variable length by warping them to match in time.

Digital Twin

A virtual model of a physical product or system used to simulate performance and optimise operations.

Demand Forecasting

Predicting future demand for products or services using machine learning to optimize inventory and marketing spend.