Linear Discriminant Analysis (LDA)
Definition
A supervised classification method that finds the linear combination of features that best separates known groups. In forensic geology it is used to classify a questioned sample into one of several source areas.
- Type
- Supervised classification statistic
- Goal
- Maximize separation between known groups
- Forensic use
- Classify a questioned geological sample by source area
Common questions
What does supervised mean in this classification context?+
The method is trained on samples whose group membership is already known, so it learns which feature combinations separate those groups before it is used to classify an unknown sample.
Why is LDA useful for geological source attribution specifically?+
Soil and mineral samples carry many overlapping chemical and physical variables. LDA combines them into the single axis that best distinguishes candidate source areas, making classification more discriminating than any one variable alone.
Related terms
- False-Positive Rate
- The proportion of authentic media files that a detection tool incorrectly classifies as manipulated or synthetic. Deepfake detectors and ELA tools both...
- Likelihood Ratio (LR)
- The ratio of two conditional probabilities: the probability of the observed evidence given the prosecution's hypothesis (same source), divided by the probability...
- Mahalanobis Distance
- A multivariate distance measure that expresses the separation between two observations in units of the population standard deviation, accounting for covariance; used...
- Principal Component Analysis (PCA)
- A multivariate statistical method used in fire debris research to reduce chromatographic data matrices to principal components that capture major variance. Used...
- Reference Population
- The set of samples representing the range of materials that could plausibly have produced the questioned sample, in the absence of the...