Principal Component Analysis (PCA)
Definition
A multivariate statistical method used in fire debris research to reduce chromatographic data matrices to principal components that capture major variance. Used as a decision-support tool in pattern recognition; cannot replace comparison sample analysis.
- Core function
- Reduces high-dimensional data (many correlated variables) to uncorrelated components that retain maximum variance.
- Common forensic use
- Visualization of multivariate samples to show clustering, separation, or position relative to reference databases.
- Application breadth
- Used across charred documents, soil comparison, fire debris, bloodstain ageing, authorship analysis, and elemental profiling.
Common questions
What does PCA do in charred document cases?+
PCA separates ink signals from the charred background by analyzing multispectral images across hundreds of wavelength bands at once. This lets forensic specialists isolate the writing on damaged documents when visible light alone isn't enough.
How does PCA help compare soil samples?+
PCA transforms dozens of correlated soil measurements (like elemental concentrations) into uncorrelated components that capture the main patterns. This visualization lets examiners plot questioned samples against known references to see how they cluster together or apart.
Can PCA replace comparison analysis in fire debris work?+
No. PCA reduces chromatographic data to show major patterns for pattern recognition and decision support. But it cannot replace the actual comparison of questioned debris against known samples, which remains essential.
Related terms
- 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...
- Pyrolysis
- The thermal decomposition of cellulose in paper at approximately 150 to 300 degrees Celsius, disrupting glycosidic bonds and releasing water and carbon...
- 4-Vinylcyclohexene (4-VCH)
- A pyrolysis product of styrene-butadiene rubber (SBR) carpet backing, appearing at m/z 108 in the GC-MS. Elevated m/z 108 in a fire...
- Alkene Triplet Pattern
- The GC-MS signature of polypropylene pyrolysis: a repeating triplet of n-alkane, 1-alkene (terminal double bond), and diene peaks at each carbon number...
- Ashing
- The terminal stage of heat damage, above approximately 400 degrees Celsius, in which carbon oxidises to carbon dioxide and only inorganic mineral...
- Background Subtraction
- The analytical practice of comparing fire debris EIPs against comparison sample EIPs from unburned substrate of the same type. Features present in...
- Bayesian Inference
- A framework for updating probability estimates in the light of new evidence; in forensic science, used to combine the prior probability of...
- Burrows's Delta
- A distance metric introduced by John Burrows in 2002 that measures how far a text's standardised word-frequency profile deviates from the mean...
- Charring
- The stage of heat damage at approximately 200 to 400 degrees Celsius at which paper cellulose converts to amorphous carbon; the sheet...
- Closed-Corpus Trap
- The error of validating a stylometric method exclusively on texts from authors whose profiles were used to build the model, inflating apparent...
- Comparison (Control) Sample
- A sample of unburned substrate collected from the same room or proximate unburned area as the fire debris sample, of the same...
Explained in these topics
- Charred and Burnt Document Recovery and DeciphermentA statistical technique applied to multispectral image cubes that separates spectral components with distinct signatures, used in charred document recovery to...
- Pyrolysis Products, Substrate Interferents and Background SubtractionA multivariate statistical method used in fire debris research to reduce chromatographic data matrices to principal components that capture major variance. Use...
- Soil Comparison Casework and Statistical InferenceA dimensionality-reduction technique that transforms correlated variables into uncorrelated principal components; used in forensic soil comparison to visualise...
- Spectroscopic Methods for Bloodstain AgeingStatistical method that reduces high-dimensional spectral data to a small number of components capturing most variance. Used in spectroscopic ageing studies to...
- Statistical Comparison and the Reference Population ProblemA technique that reduces a high-dimensional data set (e.g., twenty elemental concentrations per sample) to a smaller set of uncorrelated components that retain...
- Stylometry and Statistical Distance MethodsA dimensionality-reduction technique that projects a high-dimensional feature space onto a small number of axes capturing maximum variance. Used to visualise c...