What is earth mover's distance?
A measure of the difference between two probability distributions
Stands for: EMD
earth mover's distance explained in plain English
The earth mover's distance is a way to compare how similar or different two sets of data are, by calculating the amount of 'work' needed to transform one set into the other
Analogy
Imagine you have two piles of dirt, and you want to make them look the same. The earth mover's distance is like calculating how much dirt you need to move, and how far you need to move it, to make the two piles identical
Example
For example, it can be used to compare the distribution of pixels in two images, to determine if they are similar or not
How is earth mover's distance used?
The earth mover's distance is used in machine learning and computer vision to compare images, detect changes, and measure the similarity between data distributions
Common misconceptions about earth mover's distance
A common misconception is that the earth mover's distance is only used for image comparison, when in fact it can be applied to any type of data distribution
History
The earth mover's distance was first introduced in the 1990s, and has since been widely used in various fields, including computer science, engineering, and statistics
People also read
- A/B testing
A method of comparing two versions of a product or service to determine which one performs better
- ablation
A technique used to remove or disable parts of a machine learning model to understand their importance
- accuracy
The degree to which a model's predictions match the actual outcomes
- activation function
A mathematical function that introduces non-linearity into a neural network model
- active learning
A machine learning approach where the model actively selects the most informative data to learn from
- adaptation
The process of adjusting to new or changing conditions
- agglomerative clustering
A type of hierarchical clustering that groups similar data points together
- anomaly detection
The process of identifying data points that do not conform to expected patterns or behaviors
- area under the PR curve
A measure of a model's performance in classification tasks
- area under the ROC curve
A measure of a model's ability to distinguish between positive and negative classes